Search results for: HRV metrics

Authors: Andy Ho

Abstract:

Breast cancer is a common cancer in Australia. Early diagnosis is crucial for improving patient outcomes, as later-stage detection correlates with poorer prognoses. While multiparametric MRI offers superior sensitivity in detecting invasive and high-grade breast cancers compared to conventional mammography, its extended scan duration and high costs limit widespread application. As a result, full protocol MRI screening is typically reserved for patients at elevated risk. Recent advancements in imaging technology have facilitated the development of Abbreviated MRI protocols, which dramatically reduce scan times (<10 minutes compared to >30 minutes for full protocol). The potential for Abbreviated MRI to offer a more time- and cost-efficient alternative has implications for improving patient accessibility, reducing appointment durations, and enhancing compliance—especially relevant for individuals requiring regular annual screening over several decades. The purpose of this study is to assess the diagnostic efficacy of Abbreviated MRI for breast cancer screening among high-risk patients at the Royal Prince Alfred Hospital (RPA). This study aims to determine the sensitivity, specificity, and inter-reader variability of Abbreviated MRI protocols when interpreted by subspecialty-trained Breast Radiologists. A systematic review of the RPA’s electronic Picture Archive and Communication System identified high-risk patients, defined by Australian ‘Medicare Benefits Schedule’ criteria, who underwent Breast MRI from 2021 to 2022. Eligible participants included asymptomatic patients under 50 years old and referred by the High-Risk Clinic due to a high-risk genetic profile or relevant familial history. The MRIs were anonymized, randomized, and interpreted by four Breast Radiologists, each independently completing standardized proforma evaluations. Radiological findings were compared against histopathology as the gold standard or follow-up imaging if biopsies were unavailable. Statistical metrics, including sensitivity, specificity, and inter-reader variability, were assessed. The Fleiss-Kappa analysis demonstrated a fair inter-reader agreement (kappa = 0.25; 95% CI: 0.19–0.32; p < 0.0001). The sensitivity for detecting malignancies was 0.72, with a specificity of 0.92. For benign lesions, sensitivity and specificity were 0.844 and 0.73, respectively. These findings underline the potential of Abbreviated MRI as a reliable screening tool for malignancies with significant specificity, though reduced sensitivity highlights the importance of robust radiologist training and consistent evaluation standards. Abbreviated MRI protocols exhibit promise as a viable screening option for high-risk patients, combining reduced scan times and acceptable diagnostic accuracy. Further work to refine interpretation practices and optimize training is essential to maximize the protocol’s utility in routine clinical screening and facilitate broader accessibility.

Keywords: abbreviated, breast, cancer, MRI

Procedia PDF Downloads 10

Authors: Jiaying Zhang, Xin Mu, Chang Ni, Jeff L. Zhang

Abstract:

Magnetic resonance elastography (MRE) non-invasively assesses tissue elastic properties, such as shear modulus, by measuring tissue’s displacement in response to mechanical waves. The estimated metrics on tissue elasticity or stiffness have been shown to be valuable for monitoring physiologic or pathophysiologic status of tissue, such as a tumor or fatty liver. To quantify tissue shear modulus from MRE-acquired displacements (essentially an inverse problem), multiple approaches have been proposed, including Local Frequency Estimation (LFE) and Direct Inversion (DI). However, one common problem with these methods is that the estimates are severely noise-sensitive due to either the inverse-problem nature or noise propagation in the pixel-by-pixel process. With the advent of deep learning (DL) and its promise in solving inverse problems, a few groups in the field of MRE have explored the feasibility of using DL methods for quantifying shear modulus from MRE data. Most of the groups chose to use real MRE data for DL model training and to cut training images into smaller patches, which enriches feature characteristics of training data but inevitably increases computation time and results in outcomes with patched patterns. In this study, simulated wave images generated by Finite Differential Time Domain (FDTD) simulation are used for network training, and U-Net is used to extract features from each training image without cutting it into patches. The use of simulated data for model training has the flexibility of customizing training datasets to match specific applications. The proposed method aimed to estimate tissue shear modulus from MRE data with high robustness to noise and high model-training efficiency. Specifically, a set of 3000 maps of shear modulus (with a range of 1 kPa to 15 kPa) containing randomly positioned objects were simulated, and their corresponding wave images were generated. The two types of data were fed into the training of a U-Net model as its output and input, respectively. For an independently simulated set of 1000 images, the performance of the proposed method against DI and LFE was compared by the relative errors (root mean square error or RMSE divided by averaged shear modulus) between the true shear modulus map and the estimated ones. The results showed that the estimated shear modulus by the proposed method achieved a relative error of 4.91%±0.66%, substantially lower than 78.20%±1.11% by LFE. Using simulated data, the proposed method significantly outperformed LFE and DI in resilience to increasing noise levels and in resolving fine changes of shear modulus. The feasibility of the proposed method was also tested on MRE data acquired from phantoms and from human calf muscles, resulting in maps of shear modulus with low noise. In future work, the method’s performance on phantom and its repeatability on human data will be tested in a more quantitative manner. In conclusion, the proposed method showed much promise in quantifying tissue shear modulus from MRE with high robustness and efficiency.

Keywords: deep learning, magnetic resonance elastography, magnetic resonance imaging, shear modulus estimation

Procedia PDF Downloads 68

Authors: Ribeiro A., Vilarinho C., Luisa A., Carvalho J

Abstract:

The pig meat industry generates large volumes of by- and co-products like blood, bones, skin, trimmings, organs, viscera, and skulls, among others, during slaughtering and meat processing and must be treated and disposed of ecologically. The yield of these by-products has been reported to account for about 10% to 15% of the value of the live animal in developed countries, although animal by-products account for about two-thirds of the animal after slaughter. It was selected for further valorization of the principal wastes produced throughout the value chain of pig meat production: Pig Manure, Pig Bones, Fats, Skins, Pig Hair, Wastewater, Wastewater sludges, and other animal subproducts type III. According to the potential valorization options, these wastes will be converted into Biomethane, Fertilizers (phosphorus and digestate), Hydroxyapatite, and protein hydrolysates (Keratin and Collagen). This work includes comprehensive technical and economic analyses (TEA) for each valorization route or applied technology. Metrics such as Net Present Value (NPV), Internal Rate of Return (IRR), and payback periods were used to evaluate economic feasibility. From this analysis, it can be concluded that, for Biogas Production, the scenarios using pig manure, wastewater sludges and mixed grass and leguminous wastes presented a remarkably high economic feasibility. Scenarios showed high economic feasibility with a positive payback period, NPV, and IRR. The optimal scenario combining pig manure with mixed grass and leguminous wastes had a payback period of 1.2 years and produced 427,6269 m³ of biomethane annually. Regarding the Chemical Extraction of Phosphorous and Nitrogen, results proved that the process is economically unviable due to negative cash flows despite high recovery rates. The TEA of Hydrolysis and Extraction of Keratin Hydrolysates indicate that a unit processing and valorizing 10 tons of pig hair per year for the production of keratin hydrolysate has an NPV of 907,940 €, an IRR of 13.07%, and a Payback period of 5.41 years. All of these indicators suggest a highly potential project to explore in the future. On the opposite, the results of Hydrolysis and Extraction of Collagen Hydrolysates showed a process economically unviable with negative cash flows in all scenarios due to the high-fat content in raw materials. In fact, the results from the valorization of 10 tons of pig skin had a negative cash flow of 453 743,88 €. TEA results of Extraction and purification of Hydroxyapatite from Pig Bones with Pyrolysis indicate that unit processing and valorizing 10 tons of pig bones per year for the production of hydroxyapatite has an NPV of 1 274 819,00 €, an IRR of 65.43%, and a Payback period of 1,5 years over a timeline of 10 years with a discount rate of 10%. These valorization routes, circular economy and bio-refinery approach offer significant contributions to sustainable bio-based operations within the agri-food industry. This approach transforms waste into valuable resources, enhancing both environmental and economic outcomes and contributing to a more sustainable and circular bioeconomy.

Keywords: techno-economic analysis (TEA), pig meat processing wastes, circular economy, bio-refinery

Procedia PDF Downloads 15

Authors: Thi Phuong Pham

Abstract:

This paper explores the intricate relationships between Sustainable Supply Chain Management (SSCM) practices, digital transformation (DT), and enterprise performance within the context of Vietnam. Over the past two decades, there has been a paradigm shift in supply chain management, with sustainability gaining prominence due to increasing concerns about climate change, labor practices, and the environmental impact of business operations. In the ever-evolving realm of global business, sustainability and digital transformation (DT) intersecting dynamics have become pivotal catalysts for organizational success. This research investigates how integrating SSCM with DT can enhance enterprise performance, a subject of significant relevance as Vietnam undergoes rapid economic growth and digital transformation. The primary objectives of this research are twofold: (1) to examine the effects of SSCM practices on enterprise performance in three critical aspects: economic, environmental, and social performance in Vietnam and (2) to explore the mediating role of DT in this relationship. By analyzing these dynamics, the study aims to provide valuable insights for policymakers and the academic community regarding the potential benefits of aligning SSCM principles with digital technologies. To achieve these objectives, the research employs a robust mixed-method approach. The research begins with a comprehensive literature review to establish a theoretical framework that underpins the empirical analysis. Data collection was conducted through a structured survey targeting Vietnamese enterprises, with the survey instrument designed to measure SSCM practices, DT, and enterprise performance using a five-point Likert scale. The reliability and validity of the survey were ensured by pre-testing with industry practitioners and refining the questionnaire based on their feedback. For data analysis, structural equation modeling (SEM) was employed to quantify the direct effects of SSCM on enterprise performance, while mediation analysis using the PROCESS Macro 4.0 in SPSS was conducted to assess the mediating role of DT. The findings reveal that SSCM practices positively influence enterprise performance by enhancing operational efficiency, reducing costs, and improving sustainability metrics. Furthermore, DT acts as a significant mediator, amplifying the positive impacts of SSCM practices through improved data management, enhanced communication, and more agile supply chain processes. These results underscore the critical role of DT in maximizing the benefits of SSCM practices, particularly in a developing economy like Vietnam. This research contributes to the existing body of knowledge by highlighting the synergistic effects of SSCM and DT on enterprise performance. It offers practical implications for businesses that enhance their sustainability and digital capabilities, providing a roadmap for integrating these two pivotal aspects to achieve competitive advantage. The study's insights can also inform governmental policies designed to foster sustainable economic growth and digital innovation in Vietnam.

Keywords: sustainable supply chain management, digital transformation, enterprise performance, Vietnam

Procedia PDF Downloads 23

Authors: Atefeh Omidkhah Kharashtomi, Rasoul Hedayat Nejad, Saeed Bakhtiyari

Abstract:

Circulation paths and indoor connection network of the building play an important role both in the daily operation of the building and during evacuation in emergency situations. The degree of legibility of the paths for navigation inside the building has a deep connection with the perceptive and cognitive system of human, and the way the surrounding environment is being perceived. Human perception of the space is based on the sensory systems in a three-dimensional environment, and non-linearly, so it is necessary to avoid reducing its representations in architectural design as a two-dimensional and linear issue. Today, the advances in the field of virtual reality (VR) technology have led to various applications, and architecture and building science can benefit greatly from these capabilities. Especially in cases where the design solution requires a detailed and complete understanding of the human perception of the environment and the behavioral response, special attention to VR technologies could be a priority. Way-finding in the indoor circulation network is a proper example for such application. Success in way-finding could be achieved if human perception of the route and the behavioral reaction have been considered in advance and reflected in the architectural design. This paper discusses the VR technology applications for the way-finding improvements in indoor engineering of the building. In a systematic review, with a database consisting of numerous studies, firstly, four categories for VR applications for circulation way-finding have been identified: 1) data collection of key parameters, 2) comparison of the effect of each parameter in virtual environment versus real world (in order to improve the design), 3) comparing experiment results in the application of different VR devices/ methods with each other or with the results of building simulation, and 4) training and planning. Since the costs of technical equipment and knowledge required to use VR tools lead to the limitation of its use for all design projects, priority buildings for the use of VR during design are introduced based on case-studies analysis. The results indicate that VR technology provides opportunities for designers to solve complex buildings design challenges in an effective and efficient manner. Then environmental parameters and the architecture of the circulation routes (indicators such as route configuration, topology, signs, structural and non-structural components, etc.) and the characteristics of each (metrics such as dimensions, proportions, color, transparency, texture, etc.) are classified for the VR way-finding experiments. Then, according to human behavior and reaction in the movement-related issues, the necessity of scenario-based and experiment design for using VR technology to improve the design and receive feedback from the test participants has been described. The parameters related to the scenario design are presented in a flowchart in the form of test design, data determination and interpretation, recording results, analysis, errors, validation and reporting. Also, the experiment environment design is discussed for equipment selection according to the scenario, parameters under study as well as creating the sense of illusion in the terms of place illusion, plausibility and illusion of body ownership.

Keywords: virtual reality (VR), way-finding, indoor, circulation, design

Procedia PDF Downloads 74

Authors: A. Kourou, A. Ioakeimidou, S. Hadjiefthymiades, V. Abramea

Abstract:

During the last decades, the task of engaging governments, communities and citizens to reduce risk and vulnerability of the populations has made variable progress. Experience has demonstrated that lack of awareness, education and preparedness may result in significant material and other losses both on the onset of the disaster. Schools play a vital role in the community and are important elements of values and culture of the society. A proper school education not only teaches children, but also is a key factor in the promotion of a safety culture into the wider community. In Greece School Earthquake Safety Initiative has been undertaken by Earthquake Planning and Protection Ogranization with specific actions (seminars, lectures, guidelines, educational material, campaigns, national or EU projects, drills etc.). The objective of this initiative is to develop disaster-resilient school communities through awareness, self-help, cooperation and education. School preparedness requires the participation of Principals, teachers, students, parents, and competent authorities. Preparation and earthquake readiness involves: a) learning what should be done before, during, and after earthquake; b) doing or preparing to do these things now, before the next earthquake; and c) developing teachers’ and students’ skills to cope efficiently in case of an earthquake. In the above given framework this paper presents the results of a survey aimed to identify the level of education and preparedness of school community in Greece. More specifically, the survey questionnaire investigates issues regarding earthquake protection actions, appropriate attitudes and behaviors during an earthquake and existence of contingency plans at elementary and secondary schools. The questionnaires were administered to Principals and teachers from different regions of the country that attend the EPPO national training project 'Earthquake Safety at Schools'. A closed-form questionnaire was developed for the survey, which contained questions regarding the following: a) knowledge of self protective actions b) existence of emergency planning at home and c) existence of emergency planning at school (hazard mitigation actions, evacuation plan, and performance of drills). Survey results revealed that a high percentage of teachers have taken the appropriate preparedness measures concerning non-structural hazards at schools, emergency school plan and simulation drills every year. In order to improve the action-planning for ongoing school disaster risk reduction, the implementation of earthquake drills, the involvement of students with disabilities and the evaluation of school emergency plans, EPPO participates in E-PreS project. The main objective of this project is to create smart tools which define, simulate and evaluate all hazards emergency steps customized to the unique district and school. The project comes up with a holistic methodology using real-time evaluation involving different categories of actors, districts, steps and metrics. The project is supported by EU Civil Protection Financial Instrument with a duration of two years. Coordinator is the Kapodistrian University of Athens and partners are from four countries; Greece, Italy, Romania and Bulgaria.

Keywords: drills, earthquake, emergency plans, E-PreS project

Procedia PDF Downloads 235

Authors: I. M. L. Nadeesha Jayaweera, Adao Alex Trindade

Abstract:

In choosing a candidate model in likelihood-based modeling via an information criterion, the practitioner is often faced with the difficult task of deciding just how far up the ranked list to look. Motivated by this pragmatic necessity, we construct an uncertainty band for a generalized (model selection) information criterion (GIC), defined as a criterion for which the limit in probability is identical to that of the normalized log-likelihood. This includes common special cases such as AIC & BIC. The method starts from the asymptotic normality of the GIC for the joint distribution of the candidate models in an independent and identically distributed (IID) data framework and proceeds by deriving the (asymptotically) exact distribution of the minimum. The calculation of an upper quantile for its distribution then involves the computation of multivariate Gaussian integrals, which is amenable to efficient implementation via the R package "mvtnorm". The performance of the methodology is tested on simulated data by checking the coverage probability of nominal upper quantiles and compared to the bootstrap. Both methods give coverages close to nominal for large samples, but the bootstrap is two orders of magnitude slower. The methodology is subsequently extended to two other commonly used model structures: regression and time series. In the regression case, we derive the corresponding asymptotically exact distribution of the minimum GIC invoking Lindeberg-Feller type conditions for triangular arrays and are thus able to similarly calculate upper quantiles for its distribution via multivariate Gaussian integration. The bootstrap once again provides a default competing procedure, and we find that similar comparison performance metrics hold as for the IID case. The time series case is complicated by far more intricate asymptotic regime for the joint distribution of the model GIC statistics. Under a Gaussian likelihood, the default in most packages, one needs to derive the limiting distribution of a normalized quadratic form for a realization from a stationary series. Under conditions on the process satisfied by ARMA models, a multivariate normal limit is once again achieved. The bootstrap can, however, be employed for its computation, whence we are once again in the multivariate Gaussian integration paradigm for upper quantile evaluation. Comparisons of this bootstrap-aided semi-exact method with the full-blown bootstrap once again reveal a similar performance but faster computation speeds. One of the most difficult problems in contemporary statistical methodological research is to be able to account for the extra variability introduced by model selection uncertainty, the so-called post-model selection inference (PMSI). We explore ways in which the GIC uncertainty band can be inverted to make inferences on the parameters. This is being attempted in the IID case by pivoting the CDF of the asymptotically exact distribution of the minimum GIC. For inference one parameter at a time and a small number of candidate models, this works well, whence the attained PMSI confidence intervals are wider than the MLE-based Wald, as expected.

Keywords: model selection inference, generalized information criteria, post model selection, Asymptotic Theory

Procedia PDF Downloads 86

Authors: David Oyewumi Oyekunle, David Preston, Florence Ibeh

Abstract:

The paper provides an overview of the impacts of the human dimension in project management and team management on projects, which is increasingly affecting the performance of organizations. Recognizing its crucial significance, the research focuses on analyzing the psychological and interpersonal dynamics within project teams. This research is highly significant in the dynamic field of project management, as it addresses important gaps and offers vital insights that align with the constantly changing demands of the profession. A case study was conducted at the University of Salford to examine how human activity affects project management and performance. The study employed a mixed methodology to gain a deeper understanding of the real-world experiences of the subjects and project teams. Data analysis procedures to address the research objectives included the deductive approach, which involves testing a clear hypothesis or theory, as well as descriptive analysis and visualization. The survey comprised a sample size of 40 participants out of 110 project management professionals, including staff and final students in the Salford Business School, using a purposeful sampling method. To mitigate bias, the study ensured diversity in the sample by including both staff and final students. A smaller sample size allowed for more in-depth analysis and a focused exploration of the research objective. Conflicts, for example, are intricate occurrences shaped by a multitude of psychological stimuli and social interactions and may have either a deterrent perspective or a positive perspective on project performance and project management productivity. The study identified conflict elements, including culture, environment, personality, attitude, individual project knowledge, team relationships, leadership, and team dynamics among team members, as crucial human activities to minimize conflict. The findings are highly significant in the dynamic field of project management, as they address important gaps and offer vital insights that align with the constantly changing demands of the profession. It provided project professionals with valuable insights that can help them create a collaborative and high-performing project environment. Uncovering human influence on project performance, effective communication, optimal team synergy, and a keen understanding of project scope are necessary for the management of projects to attain exceptional performance and efficiency. For the research to achieve the aims of this study, it was acknowledged that the productive dynamics of teams and strong group cohesiveness are crucial for effectively managing conflicts in a beneficial and forward-thinking manner. Addressing the identified human influence will contribute to a more sustainable project management approach and offer opportunities for exploration and potential contributions to both academia and practical project management.

Keywords: human dimension, project management, team dynamics, conflict resolution

Procedia PDF Downloads 105

Authors: Mona Isazad Mashinchi, Davood Roshan Sangachin, Francis J. Sullivan, Dietrich Rebholz-Schuhmann

Abstract:

Background: Decision-making processes are nowadays driven by data, data analytics and Business Intelligence (BI). BI as a software platform can provide a wide variety of capabilities such as organization memory, information integration, insight creation and presentation capabilities. Visualizing data through dashboards is one of the BI solutions (for a variety of areas) which helps managers in the decision making processes to expose the most informative information at a glance. In the healthcare domain to date, dashboard presentations are more frequently used to track performance related metrics and less frequently used to monitor those quality parameters which relate directly to patient outcomes. Providing effective and timely care for patients and improving the health outcome are highly dependent on presenting and visualizing data and information. Objective: In this research, the focus is on the presentation capabilities of BI to design a dashboard for prostate cancer (PC) data that allows better decision making for the patients, the hospital and the healthcare system related to a cancer dataset. The aim of this research is to customize a retrospective PC dataset in a dashboard interface to give a better understanding of data in the categories (risk factors, treatment approaches, disease control and side effects) which matter most to patients as well as other stakeholders. By presenting the outcome in the dashboard we address one of the major targets of a value-based health care (VBHC) delivery model which is measuring the value and presenting the outcome to different actors in HC industry (such as patients and doctors) for a better decision making. Method: For visualizing the stored data to users, three interactive dashboards based on the PC dataset have been developed (using the Tableau Software) to provide better views to the risk factors, treatment approaches, and side effects. Results: Many benefits derived from interactive graphs and tables in dashboards which helped to easily visualize and see the patients at risk, better understanding the relationship between patient's status after treatment and their initial status before treatment, or to choose better decision about treatments with fewer side effects regarding patient status and etc. Conclusions: Building a well-designed and informative dashboard is related to three important factors including; the users, goals and the data types. Dashboard's hierarchies, drilling, and graphical features can guide doctors to better navigate through information. The features of the interactive PC dashboard not only let doctors ask specific questions and filter the results based on the key performance indicators (KPI) such as: Gleason Grade, Patient's Age and Status, but may also help patients to better understand different treatment outcomes, such as side effects during the time, and have an active role in their treatment decisions. Currently, we are extending the results to the real-time interactive dashboard that users (either patients and doctors) can easily explore the data by choosing preferred attribute and data to make better near real-time decisions.

Keywords: business intelligence, dashboard, decision making, healthcare, prostate cancer, value-based healthcare

Procedia PDF Downloads 140

Authors: Jason Ceavers, Travis Thompson, Juan Torres, Ramanakumar Anam, Alan Wong, Andrei Carvalho, Shane Quo, Shawn Alonso, Moises Cintron, Ricardo C. Carrero, German Lopez, Vamsi Garimella, German Giese

Abstract:

Introduction: Advance directives (AD) are essential for patients to communicate their wishes when they are not able to. Ideally, these discussions should not occur for the first time when a patient is hospitalized with an acute life-threatening illness. There is a large number of patients who do not have clearly documented ADs, resulting in the misutilization of resources and additional patient harm. This is a nationwide issue, and the Joint Commission has it as one of its national quality metrics. Presented here is a proposed protocol to increase the number of documented AD discussions in a community-based, internal medicine residency primary care clinic in South Florida. Methods: The SMART Aim for this quality improvement project is to increase documentation of AD discussions in the outpatient setting by 25% within three months in medicare patients. A survey was sent to stakeholders (clinic attendings, residents, medical assistants, front desk staff, and clinic managers), asking them for three factors they believed contributed most to the low documentation rate of AD discussions. The two most important factors were time constraints and systems issues (such as lack of a standard method to document ADs and ADs not being uploaded to the chart) which were brought up by 25% and 21.2% of the 32 survey responders, respectively. Pre-intervention data from clinic patients in 2020-2021 revealed 17.05% of patients had clear, actionable ADs documented. To address these issues, an AD pocket card was created to give to patients. One side of the card has a brief explanation of what ADs are. The other side has a column of interventions (cardiopulmonary resuscitation, mechanical ventilation, dialysis, tracheostomy, feeding tube) with boxes patients check off if they want the intervention done, do not want the intervention, do not want to discuss the topic, or need more information. These cards are to be filled out and scanned into their electronic chart to be reviewed by the resident before their appointment. The interventions that patients want more information on will be discussed by the provider. If any changes are made, the card will be re-scanned into their chart. After three months, we will chart review the patients seen in the clinic to determine how many medicare patients have a pocket card uploaded and how many have advance directives discussions documented in a progress note or annual wellness note. If there is not enough time for an AD discussion, a follow-up appointment can be scheduled for that discussion. Discussion: ADs are a crucial part of patient care, and failure to understand a patient’s wishes leads to improper utilization of resources, avoidable litigation, and patient harm. Time constraints and systems issues were identified as two major factors contributing to the lack of advance directive discussion in our community-based resident primary care clinic. Our project aims at increasing the documentation rate for ADs through a simple pocket card intervention. These are self-explanatory, easy to read and allow the patients to clearly express what interventions they desire or what they want to discuss further with their physician.

Keywords: advance directives, community-based, pocket card, primary care clinic

Procedia PDF Downloads 164

Authors: Bill Wason

Abstract:

143 Crude Palm Oil Coop mills on Sumatra Island are participating in a program to transfer land from defaulted estates to small farmers while improving the sustainability of palm production to allow for biofuel & food production. GCarbon will be working with farmers to transfer technology, fertilizer, and trees to double the yield from the current baseline of 3.5 tons to at least 7 tons of oil per ha (25 tons of fruit bunches). This will be measured via evaluation of yield comparisons between participant and non-participant farms. We will also capture methane from Palm Oil Mill Effluent (POME)throughbelt press filtering. Residues will be weighed and a formula used to estimate methane emission reductions based on methodologies developed by other researchers. GCarbon will also cover mill ponds with a non-permeable membrane and collect methane for energy or steam production. A system for accelerating methane production involving ozone and electro-flocculation will be tested to intensifymethane generation and reduce the time for wastewater treatment. A meta-analysis of research on sweet potatoes and sorghum as rotation crops will look at work in the Rio Grande do Sul, Brazil where5 ha. oftest plots of industrial sweet potato have achieved yields of 60 tons and 40 tons per ha. from 2 harvests in one year (100 MT/ha./year). Field trials will be duplicated in Bom Jesus Das Selvas, Maranhaothat will test varieties of sweet potatoes to measure yields and evaluate disease risks in a very different soil and climate of NE Brazil. Hog methane will also be captured. GCarbon Brazil, Coop Sisal, and an Australian research partner will plant several varieties of agave and use agronomic procedures to get yields of 880 MT per ha. over 5 years. They will also plant new varieties expected to get 3500 MT of biomass after 5 years (176-700 MT per ha. per year). The goal is to show that the agave can adapt to Brazil’s climate without disease problems. The study will include a field visit to growing sites in Australia where agave is being grown commercially for biofuels production. Researchers will measure the biomass per hectare at various stages in the growing cycle, sugar content at harvest, and other metrics to confirm the yield of sugar per ha. is up to 10 times greater than sugar cane. The study will look at sequestration rates from measuring soil carbon and root accumulation in various plots in Australia to confirm carbon sequestered from 5 years of production. The agave developer estimates that 60-80 MT of sequestration per ha. per year occurs from agave. The three study efforts in 3 different countries will define a feedstock pathway for jet fuel that involves very high yield crops that can produce 2 to 10 times more biomass than current assumptions. This cost-effective and less land intensive strategy will meet global jet fuel demand and produce huge quantities of food for net zero aviation and feeding 9-10 billion people by 2050

Keywords: zero emission SAF, methane capture, food-fuel integrated refining, new crops for SAF

Procedia PDF Downloads 103

Authors: Raptis Sotirios

Abstract:

Health and social care (HSc) services planning and scheduling are facing unprecedented challenges due to the pandemic pressure and also suffer from unplanned spending that is negatively impacted by the global financial crisis. Data-driven can help to improve policies, plan and design services provision schedules using algorithms assist healthcare managers’ to face unexpected demands using fewer resources. The paper discusses services packing using statistical significance tests and machine learning (ML) to evaluate demands similarity and coupling. This is achieved by predicting the range of the demand (class) using ML methods such as CART, random forests (RF), and logistic regression (LGR). The significance tests Chi-Squared test and Student test are used on data over a 39 years span for which HSc services data exist for services delivered in Scotland. The demands are probabilistically associated through statistical hypotheses that assume that the target service’s demands are statistically dependent on other demands as a NULL hypothesis. This linkage can be confirmed or not by the data. Complementarily, ML methods are used to linearly predict the above target demands from the statistically found associations and extend the linear dependence of the target’s demand to independent demands forming, thus groups of services. Statistical tests confirm ML couplings making the prediction also statistically meaningful and prove that a target service can be matched reliably to other services, and ML shows these indicated relationships can also be linear ones. Zero paddings were used for missing years records and illustrated better such relationships both for limited years and in the entire span offering long term data visualizations while limited years groups explained how well patients numbers can be related in short periods or can change over time as opposed to behaviors across more years. The prediction performance of the associations is measured using Receiver Operating Characteristic(ROC) AUC and ACC metrics as well as the statistical tests, Chi-Squared and Student. Co-plots and comparison tables for RF, CART, and LGR as well as p-values and Information Exchange(IE), are provided showing the specific behavior of the ML and of the statistical tests and the behavior using different learning ratios. The impact of k-NN and cross-correlation and C-Means first groupings is also studied over limited years and the entire span. It was found that CART was generally behind RF and LGR, but in some interesting cases, LGR reached an AUC=0 falling below CART, while the ACC was as high as 0.912, showing that ML methods can be confused padding or by data irregularities or outliers. On average, 3 linear predictors were sufficient, LGR was found competing RF well, and CART followed with the same performance at higher learning ratios. Services were packed only if when significance level(p-value) of their association coefficient was more than 0.05. Social factors relationships were observed between home care services and treatment of old people, birth weights, alcoholism, drug abuse, and emergency admissions. The work found that different HSc services can be well packed as plans of limited years, across various services sectors, learning configurations, as confirmed using statistical hypotheses.

Keywords: class, cohorts, data frames, grouping, prediction, prob-ability, services

Procedia PDF Downloads 231

Authors: Óscar Muñoz Garrigós, José Manuel Bernabeu Aubán

Abstract:

Time-sensitive services are the base of the cloud services industry. Keeping low service saturation is essential for controlling response time. All auto-scalable services make use of reactive auto-scaling. However, reactive auto-scaling has few in-depth studies. This presentation shows a model for reactive auto-scaling methodologies with a MAPE-k architecture. Queuing theory can compute different properties of static services but lacks some parameters related to the transition between models. Our model uses queuing theory parameters to relate the transition between models. It associates MAPE-k related times, the sampling frequency, the cooldown period, the number of requests that an instance can handle per unit of time, the number of incoming requests at a time instant, and a function that describes the acceleration in the service's ability to handle more requests. This model is later used as a solution to horizontally auto-scale time-sensitive services composed of microservices, reevaluating the model’s parameters periodically to allocate resources. The solution requires limiting the acceleration of the growth in the number of incoming requests to keep a constrained response time. Business benefits determine such limits. The solution can add a dynamic number of instances and remains valid under different system sizes. The study includes performance recommendations to improve results according to the incoming load shape and business benefits. The exposed methodology is tested in a simulation. The simulator contains a load generator and a service composed of two microservices, where the frontend microservice depends on a backend microservice with a 1:1 request relation ratio. A common request takes 2.3 seconds to be computed by the service and is discarded if it takes more than 7 seconds. Both microservices contain a load balancer that assigns requests to the less loaded instance and preemptively discards requests if they are not finished in time to prevent resource saturation. When load decreases, instances with lower load are kept in the backlog where no more requests are assigned. If the load grows and an instance in the backlog is required, it returns to the running state, but if it finishes the computation of all requests and is no longer required, it is permanently deallocated. A few load patterns are required to represent the worst-case scenario for reactive systems: the following scenarios test response times, resource consumption and business costs. The first scenario is a burst-load scenario. All methodologies will discard requests if the rapidness of the burst is high enough. This scenario focuses on the number of discarded requests and the variance of the response time. The second scenario contains sudden load drops followed by bursts to observe how the methodology behaves when releasing resources that are lately required. The third scenario contains diverse growth accelerations in the number of incoming requests to observe how approaches that add a different number of instances can handle the load with less business cost. The exposed methodology is compared against a multiple threshold CPU methodology allocating/deallocating 10 or 20 instances, outperforming the competitor in all studied metrics.

Keywords: reactive auto-scaling, auto-scaling, microservices, cloud computing

Procedia PDF Downloads 93

Authors: Bitewulign Mekonnen

Abstract:

Context: This scientific paper focuses on the use of near-infrared (NIR) spectroscopy to determine glucose concentration in aqueous solutions accurately and rapidly. The study compares six different machine learning methods for predicting glucose concentration and also explores the development of a deep learning model for classifying NIR spectra. The objective is to optimize the detection model and improve the accuracy of glucose prediction. This research is important because it provides a comprehensive analysis of various machine-learning techniques for estimating aqueous glucose concentrations. Research Aim: The aim of this study is to compare and evaluate different machine-learning methods for predicting glucose concentration from NIR spectra. Additionally, the study aims to develop and assess a deep-learning model for classifying NIR spectra. Methodology: The research methodology involves the use of machine learning and deep learning techniques. Six machine learning regression models, including support vector machine regression, partial least squares regression, extra tree regression, random forest regression, extreme gradient boosting, and principal component analysis-neural network, are employed to predict glucose concentration. The NIR spectra data is randomly divided into train and test sets, and the process is repeated ten times to increase generalization ability. In addition, a convolutional neural network is developed for classifying NIR spectra. Findings: The study reveals that the SVMR, ETR, and PCA-NN models exhibit excellent performance in predicting glucose concentration, with correlation coefficients (R) > 0.99 and determination coefficients (R²)> 0.985. The deep learning model achieves high macro-averaging scores for precision, recall, and F1-measure. These findings demonstrate the effectiveness of machine learning and deep learning methods in optimizing the detection model and improving glucose prediction accuracy. Theoretical Importance: This research contributes to the field by providing a comprehensive analysis of various machine-learning techniques for estimating glucose concentrations from NIR spectra. It also explores the use of deep learning for the classification of indistinguishable NIR spectra. The findings highlight the potential of machine learning and deep learning in enhancing the prediction accuracy of glucose-relevant features. Data Collection and Analysis Procedures: The NIR spectra and corresponding references for glucose concentration are measured in increments of 20 mg/dl. The data is randomly divided into train and test sets, and the models are evaluated using regression analysis and classification metrics. The performance of each model is assessed based on correlation coefficients, determination coefficients, precision, recall, and F1-measure. Question Addressed: The study addresses the question of whether machine learning and deep learning methods can optimize the detection model and improve the accuracy of glucose prediction from NIR spectra. Conclusion: The research demonstrates that machine learning and deep learning methods can effectively predict glucose concentration from NIR spectra. The SVMR, ETR, and PCA-NN models exhibit superior performance, while the deep learning model achieves high classification scores. These findings suggest that machine learning and deep learning techniques can be used to improve the prediction accuracy of glucose-relevant features. Further research is needed to explore their clinical utility in analyzing complex matrices, such as blood glucose levels.

Keywords: machine learning, signal processing, near-infrared spectroscopy, support vector machine, neural network

Procedia PDF Downloads 94

Authors: Siti Zabedah Saidin, Aidi Ahmi, Azharudin Ali, Wan Norhayati Wan Ahmad

Abstract:

In today's increasingly complex and interconnected world, the concept of sustainability has transcended mere corporate social responsibility, evolving into a fundamental driver of organizational behaviour and disclosure. This content analysis study delves into the Malaysian federal-level statutory bodies’ annual report for the year 2021, aiming to elucidate the extent of sustainability disclosures within the non-financial sections of these reports. The escalating global emphasis on sustainability has prompted organizations to embrace transparency as a means to demonstrate their commitment to environmental, social, and governance (ESG) considerations. Voluntary sustainability disclosure has emerged as a crucial channel through which organizations communicate their efforts, initiatives, and impacts in these areas, thereby fostering trust and accountability with stakeholders. The study aims to identify and examine the types of sustainability information disclosed voluntarily by the federal-level statutory bodies, concentrating on the non-financial sections of the annual reports. To achieve this, the study adopts a simplified disclosure index, a pragmatic tool that quantifies the extent of sustainability reporting in a standardized manner. Using convenience sampling, the study selects a sample of annual reports from the federal-level statutory bodies in Malaysia, as provided on their respective websites. The content analysis is centred on the non-financial sections of these reports, allowing for an in-depth exploration of sustainability disclosures. The findings of the study present the extent to which Malaysian federal-level statutory bodies embrace sustainability reporting. Through thorough content analysis, the study uncovered diverse dimensions of sustainability information, encompassing environmental impact assessments, social engagement endeavours, and governance frameworks. This reveals a deliberate effort by these bodies to encapsulate their holistic organizational contributions and challenges, transcending traditional financial metrics. This research contributes to the existing literature by providing insights into the evolving landscape of sustainability disclosure practices among Malaysian federal-level statutory bodies. The findings underline the proactive nature of these bodies in voluntarily sharing sustainability-related information, reflecting their recognition of the interconnectedness between organizational success and societal well-being. Furthermore, the study underscores the potential influence of regulatory guidelines and societal expectations in shaping the extent and nature of voluntary sustainability disclosures. Organizations are not merely responding to regulatory mandates but are actively aligning with global sustainability goals and stakeholder expectations. As organizations continue to navigate the intricate web of stakeholder expectations and sustainability imperatives, this study enriches the discourse surrounding transparency and sustainability reporting. The analysis emphasizes the important role of non-financial disclosures in portraying a holistic organizational narrative. In an era where stakeholders demand accountability, and the interconnectedness of global challenges necessitates collaborative action, the voluntary disclosure of sustainability information stands as a testament to the commitment of Malaysian federal-level statutory bodies in shaping a more sustainable future.

Keywords: voluntary disclosure, sustainability information, annual report, federal-level statutory body

Procedia PDF Downloads 64

Authors: B. R. Campomanes-Alvarez, P. Quiros, B. Fernandez

Abstract:

Automatic Speech Recognition (ASR) is a machine-based process of decoding and transcribing oral speech. A typical ASR system receives acoustic input from a speaker or an audio file, analyzes it using algorithms, and produces an output in the form of a text. Some speech recognition systems use Hidden Markov Models (HMMs) to deal with the temporal variability of speech and Gaussian Mixture Models (GMMs) to determine how well each state of each HMM fits a short window of frames of coefficients that represents the acoustic input. Another way to evaluate the fit is to use a feed-forward neural network that takes several frames of coefficients as input and produces posterior probabilities over HMM states as output. Deep neural networks (DNNs) that have many hidden layers and are trained using new methods have been shown to outperform GMMs on a variety of speech recognition systems. Acoustic models for state-of-the-art ASR systems are usually training on massive amounts of data. However, audio files with their corresponding transcriptions can be difficult to obtain, especially in the Spanish language. Hence, in the case of these low-resource scenarios, building an ASR model is considered as a complex task due to the lack of labeled data, resulting in an under-trained system. Semi-supervised learning approaches arise as necessary tasks given the high cost of transcribing audio data. The main goal of this proposal is to develop a procedure based on acoustic semi-supervised learning for Spanish ASR systems by using DNNs. This semi-supervised learning approach consists of: (a) Training a seed ASR model with a DNN using a set of audios and their respective transcriptions. A DNN with a one-hidden-layer network was initialized; increasing the number of hidden layers in training, to a five. A refinement, which consisted of the weight matrix plus bias term and a Stochastic Gradient Descent (SGD) training were also performed. The objective function was the cross-entropy criterion. (b) Decoding/testing a set of unlabeled data with the obtained seed model. (c) Selecting a suitable subset of the validated data to retrain the seed model, thereby improving its performance on the target test set. To choose the most precise transcriptions, three confidence scores or metrics, regarding the lattice concept (based on the graph cost, the acoustic cost and a combination of both), was performed as selection technique. The performance of the ASR system will be calculated by means of the Word Error Rate (WER). The test dataset was renewed in order to extract the new transcriptions added to the training dataset. Some experiments were carried out in order to select the best ASR results. A comparison between a GMM-based model without retraining and the DNN proposed system was also made under the same conditions. Results showed that the semi-supervised ASR-model based on DNNs outperformed the GMM-model, in terms of WER, in all tested cases. The best result obtained an improvement of 6% relative WER. Hence, these promising results suggest that the proposed technique could be suitable for building ASR models in low-resource environments.

Keywords: automatic speech recognition, deep neural networks, machine learning, semi-supervised learning

Procedia PDF Downloads 339

Authors: Inna Kornienko, Svetlana Guryeva, Anatoly Shekhter, Elena Petersen

Abstract:

Transplantation is an effective treatment option for patients suffering from different end-stage diseases. However, it is plagued by a constant shortage of donor organs and the subsequent need of a lifelong immunosuppressive therapy for the patient. Currently some researchers look towards using of pig organs to replace human organs for transplantation since the matrix derived from porcine organs is a convenient substitute for the human matrix. As an initial step to create a new ex vivo tissue engineered model, optimized protocols have been created to obtain organ-specific acellular matrices and evaluated their potential as tissue engineered scaffolds for culture of normal cells and tumor cell lines. These protocols include decellularization by perfusion in a bioreactor system and immersion-agitation on an orbital shaker with use of various detergents (SDS, Triton X-100) and freezing. Complete decellularization – in terms of residual DNA amount – is an important predictor of probability of immune rejection of materials of natural origin. However, the signs of cellular material may still remain within the matrix even after harsh decellularization protocols. In this regard, the matrices obtained from tissues of low-immunogenic pigs with α3Galactosyl-tranferase gene knock out (GalT-KO) may be a promising alternative to native animal sources. The research included a study of induced effect of frozen and fresh fragments of GalT-KO skin on healing of full-thickness plane wounds in 80 rats. Commercially available wound dressings (Ksenoderm, Hyamatrix and Alloderm) as well as allogenic skin were used as a positive control and untreated wounds were analyzed as a negative control. The results were evaluated on the 4th day after grafting, which corresponds to the time of start of normal wound epithelization. It has been shown that a non-specific immune response in models treated with GalT-Ko pig skin was milder than in all the control groups. Research has been performed to measure technical skin characteristics: stiffness and elasticity properties, corneometry, tevametry, and cutometry. These metrics enabled the evaluation of hydratation level, corneous layer husking level, as well as skin elasticity and micro- and macro-landscape. These preliminary data may contribute to development of personalized transplantable organs from GalT-Ko pigs with significantly limited potential of immune rejection. By applying growth factors to a decellularized skin sample it is possible to achieve various regenerative effects based on the particular situation. In this particular research BMP2 and Heparin-binding EGF-like growth factor have been used. Ideally, a bioengineered organ must be biocompatible, non-immunogenic and support cell growth. Porcine organs are attractive for xenotransplantation if severe immunologic concerns can be bypassed. The results indicate that genetically modified pig tissues with knock-outed α3Galactosyl-tranferase gene may be used for production of low-immunogenic matrix suitable for transplantation.

Keywords: decellularization, low-immunogenic, matrix, scaffolds, transplants

Procedia PDF Downloads 275

Authors: Andrew Laming, John Hattie, Mark Wilson

Abstract:

Nine Queensland Independent high schools provided deidentified student-matched ATAR and NAPLAN data for all 1217 ATAR graduates since 2020 who also sat NAPLAN at the school. Graduating cohorts from the nine schools contained a mean 100 ATAR graduates with previous NAPLAN data from their school. Excluded were vocational students (mean=27) and any ATAR graduates without NAPLAN data (mean=20). Based on Index of Community Socio-Educational Access (ICSEA) prediction, all schools had larger that predicted proportions of their students graduating with ATARs. There were an additional 173 students not releasing their ATARs to their school (14%), requiring this data to be inferred by schools. Gain was established by first converting each student’s strongest NAPLAN domain to a statewide percentile, then subtracting this result from final ATAR. The resulting ‘percentile shift’ was corrected for plausible ATAR participation at each NAPLAN level. Strongest NAPLAN domain had the highest correlation with ATAR (R2=0.58). RESULTS School mean NAPLAN scores fitted ICSEA closely (R2=0.97). Schools achieved a mean cohort gain of two ATAR rankings, but only 66% of students gained. This ranged from 46% of top-NAPLAN decile students gaining, rising to 75% achieving gains outside the top decile. The 54% of top-decile students whose ATAR fell short of prediction lost a mean 4.0 percentiles (or 6.2 percentiles prior to correction for regression to the mean). 71% of students in smaller schools gained, compared to 63% in larger schools. NAPLAN variability in each of the 13 ICSEA1100 cohorts was 17%, with both intra-school and inter-school variation of these values extremely low (0.3% to 1.8%). Mean ATAR change between years in each school was just 1.1 ATAR ranks. This suggests consecutive school cohorts and ICSEA-similar schools share very similar distributions and outcomes over time. Quantile analysis of the NAPLAN/ATAR revealed heteroscedasticity, but splines offered little additional benefit over simple linear regression. The NAPLAN/ATAR R2 was 0.33. DISCUSSION Standardised data like NAPLAN and ATAR offer educators a simple no-cost progression metric to analyse performance in conjunction with their internal test results. Change is expressed in percentiles, or ATAR shift per student, which is layperson intuitive. Findings may also reduce ATAR/vocational stream mismatch, reveal proportions of cohorts meeting or falling short of expectation and demonstrate by how much. Finally, ‘crashed’ ATARs well below expectation are revealed, which schools can reasonably work to minimise. The percentile shift method is neither value-add nor a growth percentile. In the absence of exit NAPLAN testing, this metric is unable to discriminate academic gain from legitimate ATAR-maximizing strategies. But by controlling for ICSEA, ATAR proportion variation and student mobility, it uncovers progression to ATAR metrics which are not currently publicly available. However achieved, ATAR maximisation is a sought-after private good. So long as standardised nationwide data is available, this analysis offers useful analytics for educators and reasonable predictivity when counselling subsequent cohorts about their ATAR prospects.  

Keywords: NAPLAN, ATAR, analytics, measurement, gain, performance, data, percentile, value-added, high school, numeracy, reading comprehension, variability, regression to the mean

Procedia PDF Downloads 68

Authors: Pinzhe Zhao

Abstract:

This study focuses on predicting university students' cheating tendencies using psychological traits and machine learning techniques. Academic dishonesty is a significant issue that compromises the integrity and fairness of educational institutions. While much research has been dedicated to detecting cheating behaviors after they have occurred, there is limited work on predicting such tendencies before they manifest. The aim of this research is to develop a model that can identify students who are at higher risk of engaging in academic misconduct, allowing for earlier interventions to prevent such behavior. Psychological factors are known to influence students' likelihood of cheating. Research shows that traits such as test anxiety, moral reasoning, self-efficacy, and achievement motivation are strongly linked to academic dishonesty. High levels of anxiety may lead students to cheat as a way to cope with pressure. Those with lower self-efficacy are less confident in their academic abilities, which can push them toward dishonest behaviors to secure better outcomes. Students with weaker moral judgment may also justify cheating more easily, believing it to be less wrong under certain conditions. Achievement motivation also plays a role, as students driven primarily by external rewards, such as grades, are more likely to cheat compared to those motivated by intrinsic learning goals. In this study, data on students’ psychological traits is collected through validated assessments, including scales for anxiety, moral reasoning, self-efficacy, and motivation. Additional data on academic performance, attendance, and engagement in class are also gathered to create a more comprehensive profile. Using machine learning algorithms such as Random Forest, Support Vector Machines (SVM), and Long Short-Term Memory (LSTM) networks, the research builds models that can predict students’ cheating tendencies. These models are trained and evaluated using metrics like accuracy, precision, recall, and F1 scores to ensure they provide reliable predictions. The findings demonstrate that combining psychological traits with machine learning provides a powerful method for identifying students at risk of cheating. This approach allows for early detection and intervention, enabling educational institutions to take proactive steps in promoting academic integrity. The predictive model can be used to inform targeted interventions, such as counseling for students with high test anxiety or workshops aimed at strengthening moral reasoning. By addressing the underlying factors that contribute to cheating behavior, educational institutions can reduce the occurrence of academic dishonesty and foster a culture of integrity. In conclusion, this research contributes to the growing body of literature on predictive analytics in education. It offers a approach by integrating psychological assessments with machine learning to predict cheating tendencies. This method has the potential to significantly improve how academic institutions address academic dishonesty, shifting the focus from punishment after the fact to prevention before it occurs. By identifying high-risk students and providing them with the necessary support, educators can help maintain the fairness and integrity of the academic environment.

Keywords: academic dishonesty, cheating prediction, intervention strategies, machine learning, psychological traits, academic integrity

Procedia PDF Downloads 20

Authors: Md. Abu Sufian, Robiqul Islam, Imam Hossain Shajid, Mahesh Hanumanthu, Jarasree Varadarajan, Md. Sipon Miah, Mingbo Niu

Abstract:

Coronary Heart Disease (CHD) remains a principal cause of global morbidity and mortality, characterized by atherosclerosis—the build-up of fatty deposits inside the arteries. The study introduces an innovative methodology that leverages cloud-based platforms like AWS Live Streaming and Artificial Intelligence (AI) to early detect and prevent CHD symptoms in web applications. By employing novel simulation processes and AI algorithms, this research aims to significantly mitigate the health and societal impacts of CHD. Methodology: This study introduces a novel simulation process alongside a multi-phased model development strategy. Initially, health-related data, including heart rate variability, blood pressure, lipid profiles, and ECG readings, were collected through user interactions with web-based applications as well as API Integration. The novel simulation process involved creating synthetic datasets that mimic early-stage CHD symptoms, allowing for the refinement and training of AI algorithms under controlled conditions without compromising patient privacy. AWS Live Streaming was utilized to capture real-time health data, which was then processed and analysed using advanced AI techniques. The novel aspect of our methodology lies in the simulation of CHD symptom progression, which provides a dynamic training environment for our AI models enhancing their predictive accuracy and robustness. Model Development: it developed a machine learning model trained on both real and simulated datasets. Incorporating a variety of algorithms including neural networks and ensemble learning model to identify early signs of CHD. The model's continuous learning mechanism allows it to evolve adapting to new data inputs and improving its predictive performance over time. Results and Findings: The deployment of our model yielded promising results. In the validation phase, it achieved an accuracy of 92% in predicting early CHD symptoms surpassing existing models. The precision and recall metrics stood at 89% and 91% respectively, indicating a high level of reliability in identifying at-risk individuals. These results underscore the effectiveness of combining live data streaming with AI in the early detection of CHD. Societal Implications: The implementation of cloud-based AI for CHD symptom detection represents a significant step forward in preventive healthcare. By facilitating early intervention, this approach has the potential to reduce the incidence of CHD-related complications, decrease healthcare costs, and improve patient outcomes. Moreover, the accessibility and scalability of cloud-based solutions democratize advanced health monitoring, making it available to a broader population. This study illustrates the transformative potential of integrating technology and healthcare, setting a new standard for the early detection and management of chronic diseases.

Keywords: coronary heart disease, cloud-based ai, machine learning, novel simulation techniques, early detection, preventive healthcare

Procedia PDF Downloads 64

Authors: Shayan Mohajer Hamidi

Abstract:

Federated Learning (FL) has emerged as a promising paradigm for training machine learning models on distributed data without the need for centralized data aggregation. In the realm of wireless communications, FL has the potential to leverage the vast amounts of data generated by wireless devices to improve model performance and enable intelligent applications. However, the fairness aspect of FL in wireless communications remains largely unexplored. This abstract presents an idea for fair federated learning in wireless communications, addressing the challenges of imbalanced data distribution, privacy preservation, and resource allocation. Firstly, the proposed approach aims to tackle the issue of imbalanced data distribution in wireless networks. In typical FL scenarios, the distribution of data across wireless devices can be highly skewed, resulting in unfair model updates. To address this, we propose a weighted aggregation strategy that assigns higher importance to devices with fewer samples during the aggregation process. By incorporating fairness-aware weighting mechanisms, the proposed approach ensures that each participating device's contribution is proportional to its data distribution, thereby mitigating the impact of data imbalance on model performance. Secondly, privacy preservation is a critical concern in federated learning, especially in wireless communications where sensitive user data is involved. The proposed approach incorporates privacy-enhancing techniques, such as differential privacy, to protect user privacy during the model training process. By adding carefully calibrated noise to the gradient updates, the proposed approach ensures that the privacy of individual devices is preserved without compromising the overall model accuracy. Moreover, the approach considers the heterogeneity of devices in terms of computational capabilities and energy constraints, allowing devices to adaptively adjust the level of privacy preservation to strike a balance between privacy and utility. Thirdly, efficient resource allocation is crucial for federated learning in wireless communications, as devices operate under limited bandwidth, energy, and computational resources. The proposed approach leverages optimization techniques to allocate resources effectively among the participating devices, considering factors such as data quality, network conditions, and device capabilities. By intelligently distributing the computational load, communication bandwidth, and energy consumption, the proposed approach minimizes resource wastage and ensures a fair and efficient FL process in wireless networks. To evaluate the performance of the proposed fair federated learning approach, extensive simulations and experiments will be conducted. The experiments will involve a diverse set of wireless devices, ranging from smartphones to Internet of Things (IoT) devices, operating in various scenarios with different data distributions and network conditions. The evaluation metrics will include model accuracy, fairness measures, privacy preservation, and resource utilization. The expected outcomes of this research include improved model performance, fair allocation of resources, enhanced privacy preservation, and a better understanding of the challenges and solutions for fair federated learning in wireless communications. The proposed approach has the potential to revolutionize wireless communication systems by enabling intelligent applications while addressing fairness concerns and preserving user privacy.

Keywords: federated learning, wireless communications, fairness, imbalanced data, privacy preservation, resource allocation, differential privacy, optimization

Procedia PDF Downloads 75

Authors: Chethana V. Casiker, Jagadishakumara B., Sunil G. M., Chaithra K., M. Soubadra Devy

Abstract:

A majority of the world’s food crops depends on insects for pollination, among which bees are the most dominant taxon. Bees pollinate vegetables, fruits and oilseeds which are rich in essential micronutrients. Besides being a prerequisite for a nutritionally secure diet, agrarian economies such as India depend heavily on pollination for good yield and quality of the product. As cities all over the world expand rapidly, large tracts of green spaces are being built up. This, along with high usage of agricultural chemicals has reduced floral diversity and shrunk bee habitats. Indeed, pollinator decline is being reported from various parts of the world. Further, the FAO has reported a huge increase in the area of land under cultivation of pollinator-dependent crops. In the light of increasing demand for pollination and disappearing natural habitats, it is critical to understand whether and how urban spaces can support pollinators. To this end, this study investigates the influence of landscape and local habitat quality on bee community dynamics. To capture the dynamics of expanding cityscapes, the study employs a space for time substitution, wherein a transect along the gradient of urbanization substitutes a timeframe of increasing urbanization. This will help understand how pollinators would respond to changes induced by increasing intensity of urbanization in the future. Bengaluru, one of the fastest growing cities of Southern India, is an excellent site to study impacts associated with urbanization. With sites moving away from the Bengaluru’s centre and towards its peripheries, this study captures the changes in bee species diversity and richness along a gradient of urbanization. Bees were sampled under different land use types as well as in different types of vegetation, including plantations, croplands, fallow land, parks, lake embankments, and private gardens. The relationship between bee community metrics and key drivers such as a percentage of built-up area, land use practices, and floral resources was examined. Additionally, data collected using questionnaire interviews were used to understand people’s perceptions towards and level of dependence on pollinators. Our results showed that urban areas are capable of supporting bees. In fact, a greater diversity of bees was recorded in urban sites compared to adjoining rural areas. This suggests that bees are able to seek out patchy resources and survive in small fragments of habitat. Bee abundance and species richness correlated positively with floral abundance and richness, indicating the role of vegetation in providing forage and nesting sites which are crucial to their survival. Bee numbers were seen to decrease with increase in built-up area demonstrating that impervious surfaces could act as deterrents. Findings from this study challenge the popular notion of cities being biodiversity-bare spaces. There is indeed scope for conserving bees in urban landscapes, provided that there are city-scale planning and local initiative. Bee conservation can go hand in hand with efforts such as urban gardening and terrace farming that could help cities urbanize sustainably.

Keywords: bee, landscape ecology, urbanization, urban pollination

Procedia PDF Downloads 167

Authors: Nilki Upathissa, Torin Wirasinghe

Abstract:

This solution addresses the challenges faced by traditional bookstores and the limitations of digital media, striking a balance between the tactile experience of printed books and the convenience of modern technology. The book exchange system offers a sustainable alternative, empowering users to access a diverse range of books while promoting community engagement. The user-friendly interfaces incorporated into the book exchange system ensure a seamless and enjoyable experience for users. Intuitive features for book management, search, and messaging facilitate effortless exchanges and interactions between users. By streamlining the process, the system encourages readers to explore new books aligned with their interests, enhancing the overall reading experience. Central to the system's success is the hybrid recommendation engine, which leverages advanced technologies such as Long Short-Term Memory (LSTM) models. By analyzing user input, the engine accurately predicts genre preferences, enabling personalized book recommendations. The hybrid approach integrates multiple technologies, including user interfaces, machine learning models, and recommendation algorithms, to ensure the accuracy and diversity of the recommendations. The evaluation of the book exchange system with the hybrid recommendation engine demonstrated exceptional performance across key metrics. The high accuracy score of 0.97 highlights the system's ability to provide relevant recommendations, enhancing users' chances of discovering books that resonate with their interests. The commendable precision, recall, and F1score scores further validate the system's efficacy in offering appropriate book suggestions. Additionally, the curve classifications substantiate the system's effectiveness in distinguishing positive and negative recommendations. This metric provides confidence in the system's ability to navigate the vast landscape of book choices and deliver recommendations that align with users' preferences. Furthermore, the implementation of this book exchange system with a hybrid recommendation engine has the potential to revolutionize the way readers interact with printed books. By facilitating book exchanges and providing personalized recommendations, the system encourages a sense of community and exploration within the reading community. Moreover, the emphasis on sustainability aligns with the growing global consciousness towards eco-friendly practices. With its robust technical approach and promising evaluation results, this solution paves the way for a more inclusive, accessible, and enjoyable reading experience for book lovers worldwide. In conclusion, the developed book exchange system with a hybrid recommendation engine represents a progressive solution to the challenges faced by traditional bookstores and the limitations of digital media. By promoting sustainability, widening access to printed books, and fostering engagement with reading, this system addresses the evolving needs of book enthusiasts. The integration of user-friendly interfaces, advanced machine learning models, and recommendation algorithms ensure accurate and diverse book recommendations, enriching the reading experience for users.

Keywords: recommendation systems, hybrid recommendation systems, machine learning, data science, long short-term memory, recurrent neural network

Procedia PDF Downloads 94

Authors: Ho-Won Jung, Donghun Lee, Hyung-Jin Kim

Abstract:

Predictive analytics (data analysis) uses a subset of measurements (the features, predictor, or independent variable) to predict another measurement (the outcome, target, or dependent variable) on a single person or unit. It applies empirical methods in statistics, operations research, and machine learning to predict the future, or otherwise unknown events or outcome on a single or person or unit, based on patterns in data. Most analyses of metabolic syndrome are not predictive analytics but statistical explanatory studies that build a proposed model (theory building) and then validate metabolic syndrome predictors hypothesized (theory testing). A proposed theoretical model forms with causal hypotheses that specify how and why certain empirical phenomena occur. Predictive analytics and explanatory modeling have their own territories in analysis. However, predictive analytics can perform vital roles in explanatory studies, i.e., scientific activities such as theory building, theory testing, and relevance assessment. In the context, this study is to demonstrate how to use our predictive analytics to support theory building (i.e., hypothesis generation). For the purpose, this study utilized a big data predictive analytics platform TM based on a co-occurrence graph. The co-occurrence graph is depicted with nodes (e.g., items in a basket) and arcs (direct connections between two nodes), where items in a basket are fully connected. A cluster is a collection of fully connected items, where the specific group of items has co-occurred in several rows in a data set. Clusters can be ranked using importance metrics, such as node size (number of items), frequency, surprise (observed frequency vs. expected), among others. The size of a graph can be represented by the numbers of nodes and arcs. Since the size of a co-occurrence graph does not depend directly on the number of observations (transactions), huge amounts of transactions can be represented and processed efficiently. For a demonstration, a total of 13,254 metabolic syndrome training data is plugged into the analytics platform to generate rules (potential hypotheses). Each observation includes 31 predictors, for example, associated with sociodemographic, habits, and activities. Some are intentionally included to get predictive analytics insights on variable selection such as cancer examination, house type, and vaccination. The platform automatically generates plausible hypotheses (rules) without statistical modeling. Then the rules are validated with an external testing dataset including 4,090 observations. Results as a kind of inductive reasoning show potential hypotheses extracted as a set of association rules. Most statistical models generate just one estimated equation. On the other hand, a set of rules (many estimated equations from a statistical perspective) in this study may imply heterogeneity in a population (i.e., different subpopulations with unique features are aggregated). Next step of theory development, i.e., theory testing, statistically tests whether a proposed theoretical model is a plausible explanation of a phenomenon interested in. If hypotheses generated are tested statistically with several thousand observations, most of the variables will become significant as the p-values approach zero. Thus, theory validation needs statistical methods utilizing a part of observations such as bootstrap resampling with an appropriate sample size.

Keywords: explanatory modeling, metabolic syndrome, predictive analytics, theory building

Procedia PDF Downloads 276

Authors: Adriana Haulica

Abstract:

Powered by Machine Learning, Precise medicine is tailored by now to use genetic and molecular profiling, with the aim of optimizing the therapeutic benefits for cohorts of patients. As the majority of Machine Language algorithms come from heuristics, the outputs have contextual validity. This is not very restrictive in the sense that medicine itself is not an exact science. Meanwhile, the progress made in Molecular Biology, Bioinformatics, Computational Biology, and Precise Medicine, correlated with the huge amount of human biology data and the increase in computational power, opens new healthcare challenges. A more accurate diagnosis is needed along with real-time treatments by processing as much as possible from the available information. The purpose of this paper is to present a deeper vision for the future of Artificial Intelligence in Precise medicine. In fact, actual Machine Learning algorithms use standard mathematical knowledge, mostly Euclidian metrics and standard computation rules. The loss of information arising from the classical methods prevents obtaining 100% evidence on the diagnosis process. To overcome these problems, we introduce MEDICOMPILLS, a new architectural concept tool of information processing in Precise medicine that delivers diagnosis and therapy advice. This tool processes poly-field digital resources: global knowledge related to biomedicine in a direct or indirect manner but also technical databases, Natural Language Processing algorithms, and strong class optimization functions. As the name suggests, the heart of this tool is a compiler. The approach is completely new, tailored for omics and clinical data. Firstly, the intrinsic biological intuition is different from the well-known “a needle in a haystack” approach usually used when Machine Learning algorithms have to process differential genomic or molecular data to find biomarkers. Also, even if the input is seized from various types of data, the working engine inside the MEDICOMPILLS does not search for patterns as an integrative tool. This approach deciphers the biological meaning of input data up to the metabolic and physiologic mechanisms, based on a compiler with grammars issued from bio-algebra-inspired mathematics. It translates input data into bio-semantic units with the help of contextual information iteratively until Bio-Logical operations can be performed on the base of the “common denominator “rule. The rigorousness of MEDICOMPILLS comes from the structure of the contextual information on functions, built to be analogous to mathematical “proofs”. The major impact of this architecture is expressed by the high accuracy of the diagnosis. Detected as a multiple conditions diagnostic, constituted by some main diseases along with unhealthy biological states, this format is highly suitable for therapy proposal and disease prevention. The use of MEDICOMPILLS architecture is highly beneficial for the healthcare industry. The expectation is to generate a strategic trend in Precise medicine, making medicine more like an exact science and reducing the considerable risk of errors in diagnostics and therapies. The tool can be used by pharmaceutical laboratories for the discovery of new cures. It will also contribute to better design of clinical trials and speed them up.

Keywords: bio-semantic units, multiple conditions diagnosis, NLP, omics

Procedia PDF Downloads 70

Authors: Stephan Boehm, Julia Engel, Judith Eisser

Abstract:

During the last decade communication in the Internet transformed from a broadcast to a conversational model by supporting more interactive features, enabling user generated content and introducing social media networks. Another important trend with a significant impact on electronic commerce is a massive usage shift from desktop to mobile devices. However, a presentation of product- or service-related information accumulated on websites, micro pages or portals often remains the pivot and focal point of a customer journey. A more recent change of user behavior –especially in younger user groups and in Asia– is going along with the increasing adoption of messaging applications supporting almost real-time but asynchronous communication on mobile devices. Mobile apps of this type cannot only provide an alternative for traditional one-to-one communication on mobile devices like voice calls or short messaging service. Moreover, they can be used in mobile commerce as a new marketing and sales channel, e.g., for product promotions and direct marketing activities. This requires a new way of customer interaction compared to traditional mobile commerce activities and functionalities provided based on mobile web-sites. One option better aligned to the customer interaction in mes-saging apps are so-called chatbots. Chatbots are conversational programs or dialog systems simulating a text or voice based human interaction. They can be introduced in mobile messaging and social media apps by using rule- or artificial intelligence-based imple-mentations. In this context, a comparative analysis is conducted to examine the impact of using traditional websites or chatbots for promoting a product in an impulse purchase situation. The aim of this study is to measure the impact on the customers’ user experi-ence and emotions. The study is based on a random sample of about 60 smartphone users in the group of 20 to 30-year-olds. Participants are randomly assigned into two groups and participate in a traditional website or innovative chatbot based mobile com-merce scenario. The chatbot-based scenario is implemented by using a Wizard-of-Oz experimental approach for reasons of sim-plicity and to allow for more flexibility when simulating simple rule-based and more advanced artificial intelligence-based chatbot setups. A specific set of metrics is defined to measure and com-pare the user experience in both scenarios. It can be assumed, that users get more emotionally involved when interacting with a system simulating human communication behavior instead of browsing a mobile commerce website. For this reason, innovative face-tracking and analysis technology is used to derive feedback on the emotional status of the study participants while interacting with the website or the chatbot. This study is a work in progress. The results will provide first insights on the effects of chatbot usage on user experiences and emotions in mobile commerce environments. Based on the study findings basic requirements for a user-centered design and implementation of chatbot solutions for mobile com-merce can be derived. Moreover, first indications on situations where chatbots might be favorable in comparison to the usage of traditional website based mobile commerce can be identified.

Keywords: chatbots, emotions, mobile commerce, user experience, Wizard-of-Oz prototyping

Procedia PDF Downloads 458

Authors: Rubaiat S. Ahmed, Jemer Garrido, Sergey M. Motov

Abstract:

Clinical informatics (CI) enables the transformation of data into a systematic organization that improves the quality of care and the generation of positive health outcomes.Innovative technology through informatics that compiles accurate data on analgesic utilization in the emergency department can enhance pain management in this important clinical setting. We aim to establish a simplified reporting system through CI to examine and assess the analgesic prescribing practices in the EDthrough executing a U.S. federal grant project on opioid reduction initiatives. Queried data points of interest from a level-one trauma ED’s electronic medical records were used to create data sets and develop informational/visual reporting dashboards (on Microsoft Excel and Google Sheets) concerning analgesic usage across several pre-defined parameters and performance metrics using CI. The data was then qualitatively analyzed to evaluate ED analgesic prescribing trends by departmental clinicians and leadership. During a 12-month reporting period (Dec. 1, 2020 – Nov. 30, 2021) for the ongoing project, about 41% of all ED patient visits (N = 91,747) were for pain conditions, of which 81.6% received analgesics in the ED and at discharge (D/C). Of those treated with analgesics, 24.3% received opioids compared to 75.7% receiving opioid alternatives in the ED and at D/C, including non-pharmacological modalities. Demographics showed among patients receiving analgesics, 56.7% were aged between 18-64, 51.8% were male, 51.7% were white, and 66.2% had government funded health insurance. Ninety-one percent of all opioids prescribed were in the ED, with intravenous (IV) morphine, IV fentanyl, and morphine sulfate immediate release (MSIR) tablets accounting for 88.0% of ED dispensed opioids. With 9.3% of all opioids prescribed at D/C, MSIR was dispensed 72.1% of the time. Hydrocodone, oxycodone, and tramadol usage to only 10-15% of the time, and hydromorphone at 0%. Of opioid alternatives, non-steroidal anti-inflammatory drugs were utilized 60.3% of the time, 23.5% with local anesthetics and ultrasound-guided nerve blocks, and 7.9% with acetaminophen as the primary non-opioid drug categories prescribed by ED providers. Non-pharmacological analgesia included virtual reality and other modalities. An average of 18.5 ED opioid orders and 1.9 opioid D/C prescriptions per 102.4 daily ED patient visits was observed for the period. Compared to other specialties within our institution, 2.0% of opioid D/C prescriptions are given by ED providers, compared to the national average of 4.8%. Opioid alternatives accounted for 69.7% and 30.3% usage, versus 90.7% and 9.3% for opioids in the ED and D/C, respectively.There is a pressing need for concise, relevant, and reliable clinical data on analgesic utilization for ED providers and leadership to evaluate prescribing practices and make data-driven decisions. Basic computer software can be used to create effective visual reporting dashboards with indicators that convey relevant and timely information in an easy-to-digest manner. We accurately examined our ED's analgesic prescribing practices using CI through dashboard reporting. Such reporting tools can quickly identify key performance indicators and prioritize data to enhance pain management and promote safe prescribing practices in the emergency setting.

Keywords: clinical informatics, dashboards, emergency department, health informatics, healthcare informatics, medical informatics, opioids, pain management, technology

Procedia PDF Downloads 144

Authors: Michael Sonne Kristensen, Frank Loesche, James Foster, Elif Ozcan, Judy Edworthy

Abstract:

Auditory alarms play an integral role in intensive care nurses’ daily work. Most medical devices in the intensive care unit (ICU) are designed to produce alarm sounds in order to make nurses aware of immediate or prospective safety risks. The utilisation of sound as a carrier of crucial patient information is highly dependent on nurses’ presence - both physically and mentally. For ICU nurses, especially the ones who work with stationary alarm devices at the patient bed space, it is a challenge to display ‘appropriate’ alarm responses at all times as they have to navigate with great flexibility in a complex work environment. While being primarily responsible for a small number of allocated patients they are often required to engage with other nurses’ patients, relatives, and colleagues at different locations inside and outside the unit. This work explores the social strategies used by a team of nurses to comprehend and react to the information conveyed by the alarms in the ICU. Two main research questions guide the study: To what extent do alarms from a patient bed space reach the relevant responsible nurse by direct auditory exposure? By which means do responsible nurses get informed about their patients’ alarms when not directly exposed to the alarms? A comprehensive video-ethnographic field study was carried out to capture and evaluate alarm-related events in an ICU. The study involved close collaboration with four nurses who wore eye-level cameras and ear-level binaural audio recorders during several work shifts. At all time the entire unit was monitored by multiple video and audio recorders. From a data set of hundreds of hours of recorded material information about the nurses’ location, social interaction, and alarm exposure at any point in time was coded in a multi-channel replay-interface. The data shows that responsible nurses’ direct exposure and awareness of the alarms of their allocated patients vary significantly depending on work load, social relationships, and the location of the patient’s bed space. Distributed listening is deliberately employed by the nursing team as a social strategy to respond adequately to alarms, but the patterns of information flow prompted by alarm-related events are not uniform. Auditory Spatiotemporal Trajectory (AST) is proposed as a methodological label to designate the integration of temporal, spatial and auditory load information. As a mixed-method metrics it provides tangible evidence of how nurses’ individual alarm-related experiences differ from one another and from stationary points in the ICU. Furthermore, it is used to demonstrate how alarm-related information reaches the individual nurse through principles of social and distributed cognition, and how that information relates to the actual alarm event. Thereby it bridges a long-standing gap in the literature on medical alarm utilisation between, on the one hand, initiatives to measure objective data of the medical sound environment without consideration for any human experience, and, on the other hand, initiatives to study subjective experiences of the medical sound environment without detailed evidence of the objective characteristics of the environment.

Keywords: auditory spatiotemporal trajectory, medical alarms, social cognition, video-ethography

Procedia PDF Downloads 190

Authors: Yen Yen Tan, Lynn Lam, Cynthia Lam, Angela Koh, Edwin Seng

Abstract:

Entrepreneurial education plays a crucial role in nurturing future innovators and change-makers. Over time, significant progress has been made in refining instructional approaches to develop the necessary skills among learners effectively. Two highly valuable frameworks, Kearney's "4 Principles of Entrepreneurial Pedagogy" and Bloom's "Three Domains of Learning," serve as guiding principles in entrepreneurial education. Kearney's principles align with experiential and student-centric learning, which are crucial for cultivating an entrepreneurial mindset. The potential synergies between these frameworks hold great promise for enhancing entrepreneurial acumen among students. However, despite this potential, their integration remains largely unexplored. This study aims to bridge this gap by building upon the Business Essentials through Action (BETA) module and investigating its contributions to nurturing the entrepreneurial mindset. This study employs a quasi-experimental mixed-methods approach, combining quantitative and qualitative elements to ensure comprehensive and insightful data. A cohort of 235 students participated, with 118 enrolled in the BETA module and 117 in a traditional curriculum. Their Personal Entrepreneurial Competencies (PECs) were assessed before admission (pre-Y1) and one year into the course (post-Y1) using a comprehensive 55-item PEC questionnaire, enabling measurement of critical traits such as opportunity-seeking, persistence, and risk-taking. Rigorous computations of individual entrepreneurial competencies and overall PEC scores were performed, including a correction factor to mitigate potential self-assessment bias. The orchestration of Kearney's principles and Bloom's domains within the BETA module necessitates a granular examination. Here, qualitative revelations surface, courtesy of structured interviews aligned with contemporary research methodologies. These interviews act as a portal, ushering us into the transformative journey undertaken by students. Meanwhile, the study pivots to explore the BETA module's influence on students' entrepreneurial competencies from the vantage point of faculty members. A symphony of insights emanates from intimate focus group discussions featuring six dedicated lecturers, who share their perceptions, experiences, and reflective narratives, illuminating the profound impact of pedagogical practices embedded within the BETA module. Preliminary findings from ongoing data analysis indicate promising results, showcasing a substantial improvement in entrepreneurial skills among students participating in the BETA module. This study promises not only to elevate students' entrepreneurial competencies but also to illuminate the broader canvas of applicability for Kearney's principles and Bloom's domains. The dynamic interplay of quantitative analyses, proffering precise competency metrics, and qualitative revelations, delving into the nuanced narratives of transformative journeys, engenders a holistic understanding of this educational endeavour. Through a rigorous quasi-experimental mixed-methods approach, this research aims to establish the BETA module's effectiveness in fostering entrepreneurial acumen among students at Singapore Polytechnic, thereby contributing valuable insights to the broader discourse on educational methodologies.

Keywords: entrepreneurial education, experiential learning, pedagogical frameworks, innovative competencies

Procedia PDF Downloads 64

Authors: Tolulope Aremu

Abstract:

The key process steps to produce liquid detergent products will introduce potential defects, such as formulation, mixing, filling, and packaging, which might compromise product quality, consumer safety, and operational efficiency. Real-time identification and characterization of such defects are of prime importance for maintaining high standards and reducing waste and costs. Usually, defect detection is performed by human inspection or rule-based systems, which is very time-consuming, inconsistent, and error-prone. The present study overcomes these limitations in dealing with optimization in defect characterization within the process for making liquid detergents using Machine Learning algorithms. Performance testing of various machine learning models was carried out: Support Vector Machine, Decision Trees, Random Forest, and Convolutional Neural Network on defect detection and classification of those defects like wrong viscosity, color deviations, improper filling of a bottle, packaging anomalies. These algorithms have significantly benefited from a variety of optimization techniques, including hyperparameter tuning and ensemble learning, in order to greatly improve detection accuracy while minimizing false positives. Equipped with a rich dataset of defect types and production parameters consisting of more than 100,000 samples, our study further includes information from real-time sensor data, imaging technologies, and historic production records. The results are that optimized machine learning models significantly improve defect detection compared to traditional methods. Take, for instance, the CNNs, which run at 98% and 96% accuracy in detecting packaging anomaly detection and bottle filling inconsistency, respectively, by fine-tuning the model with real-time imaging data, through which there was a reduction in false positives of about 30%. The optimized SVM model on detecting formulation defects gave 94% in viscosity variation detection and color variation. These values of performance metrics correspond to a giant leap in defect detection accuracy compared to the usual 80% level achieved up to now by rule-based systems. Moreover, this optimization with models can hasten defect characterization, allowing for detection time to be below 15 seconds from an average of 3 minutes using manual inspections with real-time processing of data. With this, the reduction in time will be combined with a 25% reduction in production downtime because of proactive defect identification, which can save millions annually in recall and rework costs. Integrating real-time machine learning-driven monitoring drives predictive maintenance and corrective measures for a 20% improvement in overall production efficiency. Therefore, the optimization of machine learning algorithms in defect characterization optimum scalability and efficiency for liquid detergent companies gives improved operational performance to higher levels of product quality. In general, this method could be conducted in several industries within the Fast moving consumer Goods industry, which would lead to an improved quality control process.

Keywords: liquid detergent manufacturing, defect detection, machine learning, support vector machines, convolutional neural networks, defect characterization, predictive maintenance, quality control, fast-moving consumer goods

Procedia PDF Downloads 18