Search results for: machine translation (MT)

Authors: Nyeng P. Gyang

Abstract:

Even though past, current and future trends suggest that multicore and cloud computing systems are increasingly prevalent/ubiquitous, this class of parallel systems is nonetheless underutilized, in general, and barely used for research on employing parallel Delaunay triangulation for parallel surface modeling and generation, in particular. The performances, of actual/physical and virtual/cloud multicore systems/machines, at executing various algorithms, which implement various parallelization strategies of the incremental insertion technique of the Delaunay triangulation algorithm, were evaluated. T-tests were run on the data collected, in order to determine whether various performance metrics differences (including execution time, speedup and efficiency) were statistically significant. Results show that the actual machine is approximately twice faster than the virtual machine at executing the same programs for the various parallelization strategies. Results, which furnish the scalability behaviors of the various parallelization strategies, also show that some of the differences between the performances of these systems, during different runs of the algorithms on the systems, were statistically significant. A few pseudo superlinear speedup results, which were computed from the raw data collected, are not true superlinear speedup values. These pseudo superlinear speedup values, which arise as a result of one way of computing speedups, disappear and give way to asymmetric speedups, which are the accurate kind of speedups that occur in the experiments performed.

Keywords: cloud computing systems, multicore systems, parallel Delaunay triangulation, parallel surface modeling and generation

Procedia PDF Downloads 202

Authors: B. Padmaja, V. V. Rama Prasad, K. V. N. Sunitha, E. Krishna Rao Patro

Abstract:

Stress is a psychological condition that reduces the quality of sleep and affects every facet of life. Constant exposure to stress is detrimental not only for mind but also body. Nevertheless, to cope with stress, one should first identify it. This paper provides an effective method for the cognitive stress level detection by using data provided from a physical activity tracker device Fitbit. This device gathers people’s daily activities of food, weight, sleep, heart rate, and physical activities. In this paper, four major stressors like physical activities, sleep patterns, working hours and change in heart rate are used to assess the stress levels of individuals. The main motive of this system is to use machine learning approach in stress detection with the help of Smartphone sensor technology. Individually, the effect of each stressor is evaluated using logistic regression and then combined model is built and assessed using variants of ordinal logistic regression models like logit, probit and complementary log-log. Then the quality of each model is evaluated using Akaike Information Criterion (AIC) and probit is assessed as the more suitable model for our dataset. This system is experimented and evaluated in a real time environment by taking data from adults working in IT and other sectors in India. The novelty of this work lies in the fact that stress detection system should be less invasive as possible for the users.

Keywords: physical activity tracker, sleep pattern, working hours, heart rate, smartphone sensor

Procedia PDF Downloads 254

Authors: Shakila N. Senarath, Dinesh Asanka, Janaka Wijayanayake

Abstract:

The necessity of cosmetic products is arising to fulfill consumer needs of personality appearance and hygiene. A cosmetic product consists of various chemical ingredients which may help to keep the skin healthy or may lead to damages. Every chemical ingredient in a cosmetic product does not perform on every human. The most appropriate way to select a healthy cosmetic product is to identify the texture of the body first and select the most suitable product with safe ingredients. Therefore, the selection process of cosmetic products is complicated. Consumer surveys have shown most of the time, the selection process of cosmetic products is done in an improper way by consumers. From this study, a content-based system is suggested that recommends cosmetic products for the human factors. To such an extent, the skin type, gender and price range will be considered as human factors. The proposed system will be implemented by using Machine Learning. Consumer skin type, gender and price range will be taken as inputs to the system. The skin type of consumer will be derived by using the Baumann Skin Type Questionnaire, which is a value-based approach that includes several numbers of questions to derive the user’s skin type to one of the 16 skin types according to the Bauman Skin Type indicator (BSTI). Two datasets are collected for further research proceedings. The user data set was collected using a questionnaire given to the public. Those are the user dataset and the cosmetic dataset. Product details are included in the cosmetic dataset, which belongs to 5 different kinds of product categories (Moisturizer, Cleanser, Sun protector, Face Mask, Eye Cream). An alternate approach of TF-IDF (Term Frequency – Inverse Document Frequency) is applied to vectorize cosmetic ingredients in the generic cosmetic products dataset and user-preferred dataset. Using the IF-IPF vectors, each user-preferred products dataset and generic cosmetic products dataset can be represented as sparse vectors. The similarity between each user-preferred product and generic cosmetic product will be calculated using the cosine similarity method. For the recommendation process, a similarity matrix can be used. Higher the similarity, higher the match for consumer. Sorting a user column from similarity matrix in a descending order, the recommended products can be retrieved in ascending order. Even though results return a list of similar products, and since the user information has been gathered, such as gender and the price ranges for product purchasing, further optimization can be done by considering and giving weights for those parameters once after a set of recommended products for a user has been retrieved.

Keywords: content-based filtering, cosmetics, machine learning, recommendation system

Procedia PDF Downloads 132

Authors: Nathaniel Carpenter

Abstract:

Web-enabled systems for supporting youth mental health management in Australia are pioneering in their field; however, with their success, these systems are experiencing exponential growth in demand which is straining an already stretched service. Supporting youth mental is critical as the lack of support is associated with significant and lasting negative consequences. To meet this growing demand, and provide critical support, investigations are needed on evaluating and improving existing online support services. Improvements should focus on developing frameworks capable of augmenting and scaling service provisions. There are few investigations informing best-practice frameworks when implementing e-mental health support systems for youth mental health; there are fewer which implement machine learning or artificially intelligent systems to facilitate the delivering of services. This investigation will use a case study methodology to highlight the design features which are important for systems to enable young people to self-manage their mental health. The investigation will also highlight the current information system challenges, to include challenges associated with service quality, provisioning, and scaling. This work will propose methods of meeting these challenges through improved design, service augmentation and automation, service quality, and through artificially intelligent inspired solutions. The results of this study will inform a framework for supporting youth mental health with intelligent and scalable web-enabled technologies to support an ever-growing user base.

Keywords: artificial intelligence, information systems, machine learning, youth mental health

Procedia PDF Downloads 109

Authors: Farshideh Einsele, Hassan Foroosh

Abstract:

In this paper, we present a robust algorithm to recognize extracted text from grocery product images captured by mobile phone cameras. Recognition of such text is challenging since text in grocery product images varies in its size, orientation, style, illumination, and can suffer from perspective distortion. Pre-processing is performed to make the characters scale and rotation invariant. Since text degradations can not be appropriately defined using wellknown geometric transformations such as translation, rotation, affine transformation and shearing, we use the whole character black pixels as our feature vector. Classification is performed with minimum distance classifier using the maximum likelihood criterion, which delivers very promising Character Recognition Rate (CRR) of 89%. We achieve considerably higher Word Recognition Rate (WRR) of 99% when using lower level linguistic knowledge about product words during the recognition process.

Keywords: camera-based OCR, feature extraction, document, image processing, grocery products

Procedia PDF Downloads 404

Authors: Sung Uk Park, Young Su Kang, Sangmo Kang, Young Kweon Suh

Abstract:

Mineral product, waste concrete (fine aggregates), waste in the optical field, industry, and construction employ separators to separate solids and classify them according to their size. Various sorting machines are used in the industrial field such as those operating under electrical properties, centrifugal force, wind power, vibration, and magnetic force. Study on separators has been carried out to contribute to the environmental industry. In this study, we perform CFD analysis for understanding the basic mechanism of the separation of waste concrete (fine aggregate) particles from air with a machine built with a rotor with blades. In CFD, we first performed two-dimensional particle tracking for various particle sizes for the model with 1 degree, 1.5 degree, and 2 degree angle between each blade to verify the boundary conditions and the method of rotating domain method to be used in 3D. Then we developed 3D numerical model with ANSYS CFX to calculate the air flow and track the particles. We judged the capability of particle separation for given size by counting the number of particles escaping from the domain toward the exit among 10 particles issued at the inlet. We confirm that particles experience stagnant behavior near the exit of the rotating blades where the centrifugal force acting on the particles is in balance with the air drag force. It was also found that the minimum particle size that can be separated by the machine with the rotor is determined by its capability to stay at the outlet of the rotor channels.

Keywords: environmental industry, separator, CFD, fine aggregate

Procedia PDF Downloads 591

Authors: Wei Yan, Wei Dong

Abstract:

Facing the understanding problem of update and preservation of urban history space under background of rapid Chinese urbanization, so at first there is a need to dig the philosophic principles of “antithesis” and “unification” which are contained in the traditional Chinese literature known as “antithesis” and do the job of planning translation by personal understanding in order to form understanding and value systems of dialectical urban history space under the foundation of “antithesis”. Then we could put forward a “context” concept for urban history space under the foregoing basis. After that, we will take the update and preservation of Ningbo Tianyi Pavilion’s historical district as an example to discuss problems related to understanding of urban history area under the basis of Chinese tradition culture, improvement of value system, construction of urban trait space and Chinese “localization” of planning theory.

Keywords: antithesis, traditional values, city renewal and conservation, the “context” of city history space

Procedia PDF Downloads 445

Authors: John Gibbs, Benjamin Flanders, Dylan Pozorski, Weixuan Liu

Abstract:

Weight Painting–adjusting the influence a skeletal joint has on a given vertex in a character mesh–is an arduous and time con- suming part of the 3D animation pipeline. This process generally requires a trained technical animator and many hours of work to complete. Our skiNNer plug-in, which works within Autodesk’s Maya 3D animation software, uses Machine Learning and data pro- cessing techniques to create a deep neural network model that can accomplish the weight painting task in seconds rather than hours for bipedal quasi-humanoid character meshes. In order to create a properly trained network, a number of challenges were overcome, including curating an appropriately large data library, managing an arbitrary 3D mesh size, handling arbitrary skeletal architectures, accounting for extreme numeric values (most data points are near 0 or 1 for weight maps), and constructing an appropriate neural network model that can properly capture the high frequency alter- ation between high weight values (near 1.0) and low weight values (near 0.0). The arrived at neural network model is a cross between a traditional CNN, deep residual network, and fully dense network. The resultant network captures the unusually hard-edged features of a weight map matrix, and produces excellent results on many bipedal models.

Keywords: 3d animation, animation, character, rigging, skinning, weight painting, machine learning, artificial intelligence, neural network, deep neural network

Procedia PDF Downloads 269

Authors: Goudjil kamel, Boukhatem Ghania, Jlailia Djihene

Abstract:

This paper delves into the development of a sophisticated desktop application designed to calculate soil bearing capacity and predict limit pressure. Drawing from an extensive review of existing methodologies, the study meticulously examines various approaches employed in soil bearing capacity calculations, elucidating their theoretical foundations and practical applications. Furthermore, the study explores the burgeoning intersection of artificial intelligence (AI) and geotechnical engineering, underscoring the transformative potential of AI- driven solutions in enhancing predictive accuracy and efficiency.Central to the research is the utilization of cutting-edge machine learning techniques, including Artificial Neural Networks (ANN), XGBoost, and Random Forest, for predictive modeling. Through comprehensive experimentation and rigorous analysis, the efficacy and performance of each method are rigorously evaluated, with XGBoost emerging as the preeminent algorithm, showcasing superior predictive capabilities compared to its counterparts. The study culminates in a nuanced understanding of the intricate dynamics at play in geotechnical analysis, offering valuable insights into optimizing soil bearing capacity calculations and limit pressure predictions. By harnessing the power of advanced computational techniques and AI-driven algorithms, the paper presents a paradigm shift in the realm of geotechnical engineering, promising enhanced precision and reliability in civil engineering projects.

Keywords: limit pressure of soil, xgboost, random forest, bearing capacity

Procedia PDF Downloads 11

Authors: Arfan Ali Nagra, Tariq Shahzad, Meshal Alharbi, Khalid Masood Khan, Muhammad Mugees Asif, Taher M. Ghazal, Khmaies Ouahada

Abstract:

Gene selection is an essential step for the classification of microarray cancer data. Gene expression cancer data (DNA microarray) facilitates computing the robust and concurrent expression of various genes. Particle swarm optimization (PSO) requires simple operators and less number of parameters for tuning the model in gene selection. The selection of a prognostic gene with small redundancy is a great challenge for the researcher as there are a few complications in PSO based selection method. In this research, a new variant of PSO (Self-inertia weight adaptive PSO) has been proposed. In the proposed algorithm, SIW-APSO-ELM is explored to achieve gene selection prediction accuracies. This new algorithm balances the exploration capabilities of the improved inertia weight adaptive particle swarm optimization and the exploitation. The self-inertia weight adaptive particle swarm optimization (SIW-APSO) is used to search the solution. The SIW-APSO is updated with an evolutionary process in such a way that each particle iteratively improves its velocities and positions. The extreme learning machine (ELM) has been designed for the selection procedure. The proposed method has been to identify a number of genes in the cancer dataset. The classification algorithm contains ELM, K- centroid nearest neighbor (KCNN), and support vector machine (SVM) to attain high forecast accuracy as compared to the start-of-the-art methods on microarray cancer datasets that show the effectiveness of the proposed method.

Keywords: microarray cancer, improved PSO, ELM, SVM, evolutionary algorithms

Procedia PDF Downloads 81

Authors: Phillip Kearney, Constantina Lekakou, Stephen Belcher, Alessandro Sordon

Abstract:

The focus in the automotive industry is to reduce human operator and machine interaction, so manufacturing becomes more automated and safer. The aim is to lower part cost and construction time as well as defects in the parts, sometimes occurring due to the physical limitations of human operators. A move to automate the layup of reinforcement material in composites manufacturing has resulted in the use of tapes that are placed in position by a robotic deposition head, also described as Automated Fibre Placement (AFP). The process of AFP is limited with respect to the finite amount of material that can be loaded into the machine at any one time. Joining two batches of tape material together involves a splice to secure the ends of the finishing tape to the starting edge of the new tape. The splicing method of choice for the majority of prepreg applications is a hand stich method, and as the name suggests requires human input to achieve. This investigation explores three methods for automated splicing, namely, adhesive, binding and stitching. The adhesive technique uses an additional adhesive placed on the tape ends to be joined. Binding uses the binding agent that is already impregnated onto the tape through the application of heat. The stitching method is used as a baseline to compare the new splicing methods to the traditional technique currently in use. As the methods will be used within a High Speed Automated Fibre Placement (HSAFP) process, this meant the parameters of the splices have to meet certain specifications: (a) the splice must be able to endure a load of 50 N in tension applied at a rate of 1 mm/s; (b) the splice must be created in less than 6 seconds, dictated by the capacity of the tape accumulator within the system. The samples for experimentation were manufactured with controlled overlaps, alignment and splicing parameters, these were then tested in tension using a tensile testing machine. Initial analysis explored the use of the impregnated binding agent present on the tape, as in the binding splicing technique. It analysed the effect of temperature and overlap on the strength of the splice. It was found that the optimum splicing temperature was at the higher end of the activation range of the binding agent, 100 °C. The optimum overlap was found to be 25 mm; it was found that there was no improvement in bond strength from 25 mm to 30 mm overlap. The final analysis compared the different splicing methods to the baseline of a stitched bond. It was found that the addition of an adhesive was the best splicing method, achieving a maximum load of over 500 N compared to the 26 N load achieved by a stitching splice and 94 N by the binding method.

Keywords: analysis, automated fibre placement, high speed, splicing

Procedia PDF Downloads 153

Authors: Muhammad Umair, Yuri Nikolaev, Denis Artemov, Ighor Uzhinsky

Abstract:

A flexible manufacturing system is an integral part of a smart factory of industry 4.0 in which every machine is interconnected and works autonomously. Robots are in the process of replacing humans in every industrial sector. As the cyber-physical-system (CPS) and artificial intelligence (AI) are advancing, the manufacturing industry is getting more dependent on computers than human brains. This modernization has boosted the production with high quality and accuracy and shifted from classic production to smart manufacturing systems. However, material handling for such automated productions is a challenge and needs to be addressed with the best possible solution. Conventional clamping systems are designed for manual work and not suitable for highly automated production systems. Researchers and engineers are trying to find the most economical solution for loading/unloading and transportation workpieces from a warehouse to a machine shop for machining operations and back to the warehouse without human involvement. This work aims to propose an advanced multi-shape tooling solution for highly automated manufacturing systems. The currently obtained result shows that it could function well with automated guided vehicles (AGVs) and modern conveyor belts. The proposed solution is following requirements to be automation-friendly, universal for different part geometry and production operations. We used a bottom-up approach in this work, starting with studying different case scenarios and their limitations and finishing with the general solution.

Keywords: artificial intelligence, cyber physics system, Industry 4.0, material handling, smart factory, flexible manufacturing system

Procedia PDF Downloads 127

Authors: Yingisani Chabalala, Elhadi Adam, Khalid Adem Ali

Abstract:

Mapping smallholder fruit plantations using optical data is challenging due to morphological landscape heterogeneity and crop types having overlapped spectral signatures. Furthermore, cloud covers limit the use of optical sensing, especially in subtropical climates where they are persistent. This research assessed the effectiveness of Sentinel-1 (S1) and Sentinel-2 (S2) data for mapping fruit trees and co-existing land-use types by using support vector machine (SVM) and random forest (RF) classifiers independently. These classifiers were also applied to fused data from the two sensors. Feature ranks were extracted using the RF mean decrease accuracy (MDA) and forward variable selection (FVS) to identify optimal spectral windows to classify fruit trees. Based on RF MDA and FVS, the SVM classifier resulted in relatively high classification accuracy with overall accuracy (OA) = 0.91.6% and kappa coefficient = 0.91% when applied to the fused satellite data. Application of SVM to S1, S2, S2 selected variables and S1S2 fusion independently produced OA = 27.64, Kappa coefficient = 0.13%; OA= 87%, Kappa coefficient = 86.89%; OA = 69.33, Kappa coefficient = 69. %; OA = 87.01%, Kappa coefficient = 87%, respectively. Results also indicated that the optimal spectral bands for fruit tree mapping are green (B3) and SWIR_2 (B10) for S2, whereas for S1, the vertical-horizontal (VH) polarization band. Including the textural metrics from the VV channel improved crop discrimination and co-existing land use cover types. The fusion approach proved robust and well-suited for accurate smallholder fruit plantation mapping.

Keywords: smallholder agriculture, fruit trees, data fusion, precision agriculture

Procedia PDF Downloads 49

Authors: Gaurav Kumar Sinha

Abstract:

The oil refining industry, significant in its energy consumption and carbon emissions, faces increasing pressure to reduce its environmental footprint. This article explores the application of energy efficiency and sustainability analytics as crucial tools for reducing carbon emissions in oil refineries. Through a comprehensive review of current practices and technologies, this study highlights innovative analytical approaches that can significantly enhance energy efficiency. We focus on the integration of advanced data analytics, including machine learning and predictive modeling, to optimize process controls and energy use. These technologies are examined for their potential to not only lower energy consumption but also reduce greenhouse gas emissions. Additionally, the article discusses the implementation of sustainability analytics to monitor and improve environmental performance across various operational facets of oil refineries. We explore case studies where predictive analytics have successfully identified opportunities for reducing energy use and emissions, providing a template for industry-wide application. The challenges associated with deploying these analytics, such as data integration and the need for skilled personnel, are also addressed. The paper concludes with strategic recommendations for oil refineries aiming to enhance their sustainability practices through the adoption of targeted analytics. By implementing these measures, refineries can achieve significant reductions in carbon emissions, aligning with global environmental goals and regulatory requirements.

Keywords: energy efficiency, sustainability analytics, carbon emissions, oil refineries, data analytics, machine learning, predictive modeling, process optimization, greenhouse gas reduction, environmental performance

Procedia PDF Downloads 27

Authors: Sidra Jawed

Abstract:

The accessibility constructs and the body privilege discourse has been a major problem while dealing with health inequities and inaccessibility. The inherent problem in this arbitrary view of disability is that disability would never be the productive way of living. For past thirty years, disability activists have been working to differentiate ‘impairment’ from ‘disability’ and probing for more understanding of limitation imposed by society, this notion is ultimately known as the Social Model of Disability. The vulnerable population as disability community remains marginalized and seen relentlessly fighting to highlight the importance of social factors. It does not only constitute physical architectural barriers and famous blue symbol of access to the healthcare but also invisible, intangible barriers as attitudes and behaviours. Conventionally the idea of ‘disability’ has been laden with prejudiced perception amalgamating with biased attitude. Equity in contemporary setup necessitates the restructuring of organizational structure. Apparently simple, the complex interplay of disability and contemporary healthcare set up often ends up at negotiating vital components of basic healthcare needs. The role of society is indispensable when it comes to people with disability (PWD), everything from the access to healthcare to timely interventions are strongly related to the set up in place and the attitude of healthcare providers. It is vital to understand the association between assumptions and the quality of healthcare PWD receives in our global healthcare setup. Most of time the crucial physician-patient relationship with PWD is governed by the negative assumptions of the physicians. The multifaceted, troubled patient-physicians’ relationship has been neglected in past. To compound it, insufficient work has been done to explore physicians’ perspective about the disability and access to healthcare PWD have currently. This research project is directed towards physicians’ perspective on the intersection of health and access of healthcare for PWD. The principal aim of the study is to explore the perception of disability in family medicine physicians, highlighting the underpinning of medical perspective in healthcare institution. In the quest of removing barriers, the first step must be to identify the barriers and formulate a plan for future policies, involving all the stakeholders. There would be semi-structured interviews to explore themes as accessibility, medical training, construct of social model and medical model of disability, time limitations, financial constraints. The main research interest is to identify the obstacles to inclusion and marginalization continuing from the basic living necessities to wide health inequity in present society. Physicians point of view is largely missing from the research landscape and the current forum of knowledge with regards to physicians’ standpoint. This research will provide policy makers with a starting point and comprehensive background knowledge that can be a stepping stone for future researches and furthering the knowledge translation process to strengthen healthcare. Additionally, it would facilitate the process of knowledge translation between the much needed medical and disability community.

Keywords: disability, physicians, social model, accessibility

Procedia PDF Downloads 217

Authors: Serge Tshibangu, Turgay Celik, Zenzo Ncube

Abstract:

Road crashes kill nearly over a million people every year, and leave millions more injured or permanently disabled. Various annual reports reveal that the percentage of fatal crashes due to fatigue/driver falling asleep comes directly after the percentage of fatal crashes due to intoxicated drivers. This percentage is higher than the combined percentage of fatal crashes due to illegal/Un-Safe U-turn and illegal/Un-Safe reversing. Although a relatively small percentage of police reports on road accidents highlights drowsiness and fatigue, the importance of these factors is greater than we might think, hidden by the undercounting of their events. Some scenarios show that these factors are significant in accidents with killed and injured people. Thus the need for an automatic drivers fatigue detection system in order to considerably reduce the number of accidents owing to fatigue.This research approaches the drivers fatigue detection problem in an innovative way by combining cues collected from both temporal analysis of drivers’ faces and environment. Monotony in driving environment is inter-related with visual symptoms of fatigue on drivers’ faces to achieve fatigue detection. Optical and infrared (IR) sensors are used to analyse the monotony in driving environment and to detect the visual symptoms of fatigue on human face. Internal cues from drivers faces and external cues from environment are combined together using machine learning algorithms to automatically detect fatigue.

Keywords: intelligent transportation systems, bayesian networks, yawning computing, machine learning algorithms

Procedia PDF Downloads 453

Authors: Ammar Alali, Mahmoud Abughaban

Abstract:

Cost reduction and drilling optimization is the goal of many drilling operators. Historically, stuck pipe incidents were a major segment of non-productive time (NPT) associated costs. Traditionally, stuck pipe problems are part of the operations and solved post-sticking. However, the real key to savings and success is in predicting the stuck pipe incidents and avoiding the conditions leading to its occurrences. Previous attempts in stuck-pipe predictions have neglected the local geology of the problem. The proposed predictive tool utilizes geophysical data processing techniques and Machine Learning (ML) algorithms to predict drilling activities events in real-time using surface drilling data with minimum computational power. The method combines two types of analysis: (1) real-time prediction, and (2) cause analysis. Real-time prediction aggregates the input data, including historical drilling surface data, geological formation tops, and petrophysical data, from wells within the same field. The input data are then flattened per the geological formation and stacked per stuck-pipe incidents. The algorithm uses two physical methods (stacking and flattening) to filter any noise in the signature and create a robust pre-determined pilot that adheres to the local geology. Once the drilling operation starts, the Wellsite Information Transfer Standard Markup Language (WITSML) live surface data are fed into a matrix and aggregated in a similar frequency as the pre-determined signature. Then, the matrix is correlated with the pre-determined stuck-pipe signature for this field, in real-time. The correlation used is a machine learning Correlation-based Feature Selection (CFS) algorithm, which selects relevant features from the class and identifying redundant features. The correlation output is interpreted as a probability curve of stuck pipe incidents prediction in real-time. Once this probability passes a fixed-threshold defined by the user, the other component, cause analysis, alerts the user of the expected incident based on set pre-determined signatures. A set of recommendations will be provided to reduce the associated risk. The validation process involved feeding of historical drilling data as live-stream, mimicking actual drilling conditions, of an onshore oil field. Pre-determined signatures were created for three problematic geological formations in this field prior. Three wells were processed as case studies, and the stuck-pipe incidents were predicted successfully, with an accuracy of 76%. This accuracy of detection could have resulted in around 50% reduction in NPT, equivalent to 9% cost saving in comparison with offset wells. The prediction of stuck pipe problem requires a method to capture geological, geophysical and drilling data, and recognize the indicators of this issue at a field and geological formation level. This paper illustrates the efficiency and the robustness of the proposed cross-disciplinary approach in its ability to produce such signatures and predicting this NPT event.

Keywords: drilling optimization, hazard prediction, machine learning, stuck pipe

Procedia PDF Downloads 223

Authors: Ja-Keoung Koo, Kensuke Nakamura, Hyohun Kim, Dongwha Shin, Yeonseok Kim, Ji-Su Ahn, Byung-Woo Hong

Abstract:

The machine learning techniques based on a convolutional neural network (CNN) have been actively developed and successfully applied to a variety of image analysis tasks including reconstruction, noise reduction, resolution enhancement, segmentation, motion estimation, object recognition. The classical visual information processing that ranges from low level tasks to high level ones has been widely developed in the deep learning framework. It is generally considered as a challenging problem to derive visual interpretation from high dimensional imagery data. A CNN is a class of feed-forward artificial neural network that usually consists of deep layers the connections of which are established by a series of non-linear operations. The CNN architecture is known to be shift invariant due to its shared weights and translation invariance characteristics. However, it is often computationally intractable to optimize the network in particular with a large number of convolution layers due to a large number of unknowns to be optimized with respect to the training set that is generally required to be large enough to effectively generalize the model under consideration. It is also necessary to limit the size of convolution kernels due to the computational expense despite of the recent development of effective parallel processing machinery, which leads to the use of the constantly small size of the convolution kernels throughout the deep CNN architecture. However, it is often desired to consider different scales in the analysis of visual features at different layers in the network. Thus, we propose a CNN model where different sizes of the convolution kernels are applied at each layer based on the random projection. We apply random filters with varying sizes and associate the filter responses with scalar weights that correspond to the standard deviation of the random filters. We are allowed to use large number of random filters with the cost of one scalar unknown for each filter. The computational cost in the back-propagation procedure does not increase with the larger size of the filters even though the additional computational cost is required in the computation of convolution in the feed-forward procedure. The use of random kernels with varying sizes allows to effectively analyze image features at multiple scales leading to a better generalization. The robustness and effectiveness of the proposed CNN based on random kernels are demonstrated by numerical experiments where the quantitative comparison of the well-known CNN architectures and our models that simply replace the convolution kernels with the random filters is performed. The experimental results indicate that our model achieves better performance with less number of unknown weights. The proposed algorithm has a high potential in the application of a variety of visual tasks based on the CNN framework. Acknowledgement—This work was supported by the MISP (Ministry of Science and ICT), Korea, under the National Program for Excellence in SW (20170001000011001) supervised by IITP, and NRF-2014R1A2A1A11051941, NRF2017R1A2B4006023.

Keywords: deep learning, convolutional neural network, random kernel, random projection, dimensionality reduction, object recognition

Procedia PDF Downloads 284

Authors: M. Aghili, S. Tabarestani, C. Freytes, M. Shojaie, M. Cabrerizo, A. Barreto, N. Rishe, R. E. Curiel, D. Loewenstein, R. Duara, M. Adjouadi

Abstract:

A major challenge in medical studies, especially those that are longitudinal, is the problem of missing measurements which hinders the effective application of many machine learning algorithms. Furthermore, recent Alzheimer's Disease studies have focused on the delineation of Early Mild Cognitive Impairment (EMCI) and Late Mild Cognitive Impairment (LMCI) from cognitively normal controls (CN) which is essential for developing effective and early treatment methods. To address the aforementioned challenges, this paper explores the potential of using the eXtreme Gradient Boosting (XGBoost) algorithm in handling missing values in multiclass classification. We seek a generalized classification scheme where all prodromal stages of the disease are considered simultaneously in the classification and decision-making processes. Given the large number of subjects (1631) included in this study and in the presence of almost 28% missing values, we investigated the performance of XGBoost on the classification of the four classes of AD, NC, EMCI, and LMCI. Using 10-fold cross validation technique, XGBoost is shown to outperform other state-of-the-art classification algorithms by 3% in terms of accuracy and F-score. Our model achieved an accuracy of 80.52%, a precision of 80.62% and recall of 80.51%, supporting the more natural and promising multiclass classification.

Keywords: eXtreme gradient boosting, missing data, Alzheimer disease, early mild cognitive impairment, late mild cognitive impair, multiclass classification, ADNI, support vector machine, random forest

Procedia PDF Downloads 186

Authors: Amnon Shirizly, Ohad Dolev

Abstract:

In recent years, the implementation and use of Metal Additive Manufacturing (AM) parts increase. As a result, the demand for bigger parts rises along with the desire to reduce it’s the production cost. Generally, in powder bed Additive Manufacturing technology the part size is limited by the machine build volume. In order to overcome this limitation, the parts can be built in one or more machine operations and mechanically joint or weld them together. An alternative option could be a production of wrought part and built on it the AM structure (mainly to reduce costs). In both cases, the mechanical properties of the interface have to be defined and recognized. In the current study, the authors introduce guidelines on how to examine the interface between wrought alloy and powder-bed AM. The mechanical and metallurgical properties of the Ti6Al4V materials (wrought alloy and powder-bed AM) and their hybrid interface were examined. The mechanical properties gain from tensile test bars in the built direction and fracture toughness samples in various orientations. The hybrid specimens were built onto a wrought Ti6Al4V start-plate. The standard fracture toughness (CT25 samples) and hybrid tensile specimens' were heat treated and milled as a post process to final diminutions. In this Study, the mechanical tensile tests and fracture toughness properties supported by metallurgical observation will be introduced and discussed. It will show that the hybrid approach of utilizing powder bed AM onto wrought material expanding the current limitation of the future manufacturing technology.

Keywords: additive manufacturing, hybrid, fracture-toughness, powder bed

Procedia PDF Downloads 102

Authors: Yash Jain

Abstract:

The US Center for Disease Control has recommended wearing masks to slow the spread of the virus. The research uses a video feed from a camera to conduct real-time classifications of whether or not a human is correctly wearing a mask, incorrectly wearing a mask, or not wearing a mask at all. Utilizing two distinct datasets from the open-source website Kaggle, a mask detection network had been trained. The first dataset that was used to train the model was titled 'Face Mask Detection' on Kaggle, where the dataset was retrieved from and the second dataset was titled 'Face Mask Dataset, which provided the data in a (YOLO Format)' so that the TinyYoloV3 model could be trained. Based on the data from Kaggle, two machine learning models were implemented and trained: a Tiny YoloV3 Real-time model and a two-stage neural network classifier. The two-stage neural network classifier had a first step of identifying distinct faces within the image, and the second step was a classifier to detect the state of the mask on the face and whether it was worn correctly, incorrectly, or no mask at all. The TinyYoloV3 was used for the live feed as well as for a comparison standpoint against the previous two-stage classifier and was trained using the darknet neural network framework. The two-stage classifier attained a mean average precision (MAP) of 80%, while the model trained using TinyYoloV3 real-time detection had a mean average precision (MAP) of 59%. Overall, both models were able to correctly classify stages/scenarios of no mask, mask, and incorrectly worn masks.

Keywords: datasets, classifier, mask-detection, real-time, TinyYoloV3, two-stage neural network classifier

Procedia PDF Downloads 157

Authors: Lai Wenfang

Abstract:

Tthis study will describe how to use artificial intelligence (AI) technology to build the user-oriented platform for integrated archival service. The platform will be launched in 2020 by the National Archives Administration (NAA) in Taiwan. With the progression of information communication technology (ICT) the NAA has built many systems to provide archival service. In order to cope with new challenges, such as new ICT, artificial intelligence or blockchain etc. the NAA will try to use the natural language processing (NLP) and machine learning (ML) skill to build a training model and propose suggestions based on the data sent to the platform. NAA expects the platform not only can automatically inform the sending agencies’ staffs which records catalogues are against the transfer or destroy rules, but also can use the model to find the details hidden in the catalogues and suggest NAA’s staff whether the records should be or not to be, to shorten the auditing time. The platform keeps all the users’ browse trails; so that the platform can predict what kinds of archives user could be interested and recommend the search terms by visualization, moreover, inform them the new coming archives. In addition, according to the Archives Act, the NAA’s staff must spend a lot of time to mark or remove the personal data, classified data, etc. before archives provided. To upgrade the archives access service process, the platform will use some text recognition pattern to black out automatically, the staff only need to adjust the error and upload the correct one, when the platform has learned the accuracy will be getting higher. In short, the purpose of the platform is to deduct the government digital transformation and implement the vision of a service-oriented smart government.

Keywords: artificial intelligence, natural language processing, machine learning, visualization

Procedia PDF Downloads 169

Authors: Songul Cinaroglu

Abstract:

Regional planning in health has gained speed for developing countries in recent years. In Turkey, 89 different Public Hospital Unions (PHUs) were conducted based on provincial levels. In this study technical efficiencies of 89 PHUs were examined by using Data Envelopment Analysis (DEA) and machine learning techniques by dividing them into two clusters in terms of similarities of input and output indicators. Number of beds, physicians and nurses determined as input variables and number of outpatients, inpatients and surgical operations determined as output indicators. Before performing DEA, PHUs were grouped into two clusters. It is seen that the first cluster represents PHUs which have higher population, demand and service density than the others. The difference between clusters was statistically significant in terms of all study variables (p ˂ 0.001). After clustering, DEA was performed for general and for two clusters separately. It was found that 11% of PHUs were efficient in general, additionally 21% and 17% of them were efficient for the first and second clusters respectively. It is seen that PHUs, which are representing urban parts of the country and have higher population and service density, are more efficient than others. Random forest decision tree graph shows that number of inpatients is a determinative factor of efficiency of PHUs, which is a measure of service density. It is advisable for public health policy makers to use statistical learning methods in resource planning decisions to improve efficiency in health care.

Keywords: public hospital unions, efficiency, data envelopment analysis, random forest

Procedia PDF Downloads 124

Authors: Rabie Fadil, Parshuram Aarotale, Shubha Majumder, Bijay Guargain

Abstract:

Hemorrhage is the loss of blood from the circulatory system and leading cause of battlefield and postpartum related deaths. Early detection of hemorrhage remains the most effective strategy to reduce mortality rate caused by traumatic injuries. In this study, we investigated the physiological changes via non-invasive cardiac signals at rest and under different hemorrhage conditions simulated through graded lower-body negative pressure (LBNP). Simultaneous electrocardiogram (ECG), photoplethysmogram (PPG), blood pressure (BP), impedance cardiogram (ICG), and phonocardiogram (PCG) were acquired from 10 participants (age:28 ± 6 year, weight:73 ± 11 kg, height:172 ± 8 cm). The LBNP protocol consisted of applying -20, -30, -40, -50, and -60 mmHg pressure to the lower half of the body. Beat-to-beat heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean aerial pressure (MAP) were extracted from ECG and blood pressure. Systolic amplitude (SA), systolic time (ST), diastolic time (DT), and left ventricle Ejection time (LVET) were extracted from PPG during each stage. Preliminary results showed that the application of -40 mmHg i.e. moderate stage simulated hemorrhage resulted significant changes in HR (85±4 bpm vs 68 ± 5bpm, p < 0.01), ST (191 ± 10 ms vs 253 ± 31 ms, p < 0.05), LVET (350 ± 14 ms vs 479 ± 47 ms, p < 0.05) and DT (551 ± 22 ms vs 683 ± 59 ms, p < 0.05) compared to rest, while no change was observed in SA (p > 0.05) as a consequence of LBNP application. These findings demonstrated the potential of cardiac signals in detecting moderate hemorrhage. In future, we will analyze all the LBNP stages and investigate the feasibility of other physiological signals to develop a predictive machine learning model for early detection of hemorrhage.

Keywords: blood pressure, hemorrhage, lower-body negative pressure, LBNP, machine learning

Procedia PDF Downloads 166

Authors: Xi Liu, Fan Yuan

Abstract:

In the evolving landscape of physics-based machine learning (PBML), particularly within fluid dynamics and its applications in electromechanical engineering, robot vision, and robot learning, achieving precision and alignment with researchers' specific needs presents a formidable challenge. In response, this work proposes a methodology that integrates neural transformation with a modified smoothed particle hydrodynamics model for generating transformed 3D fluid simulations. This approach is useful for nanoscale science, where the unique and complex behaviors of viscoelastic medium demand accurate neurally-transformed simulations for materials understanding and manipulation. In electromechanical engineering, the method enhances the design and functionality of fluid-operated systems, particularly microfluidic devices, contributing to advancements in nanomaterial design, drug delivery systems, and more. The proposed approach also aligns with the principles of PBML, offering advantages such as multi-fluid stylization and consistent particle attribute transfer. This capability is valuable in various fields where the interaction of multiple fluid components is significant. Moreover, the application of neurally-transformed hydrodynamical models extends to manufacturing processes, such as the production of microelectromechanical systems, enhancing efficiency and cost-effectiveness. The system's ability to perform neural transfer on 3D fluid scenes using a deep learning algorithm alongside physical models further adds a layer of flexibility, allowing researchers to tailor simulations to specific needs across scientific and engineering disciplines.

Keywords: physics-based machine learning, robot vision, robot learning, hydrodynamics

Procedia PDF Downloads 63

Authors: Vedant Janapaty

Abstract:

Estuaries are wetlands where fresh water from streams mixes with salt water from the sea. Also known as “kidneys of our planet”- they are extremely productive environments that filter pollutants, absorb floods from sea level rise, and shelter a unique ecosystem. However, eutrophication and loss of native species are ailing our wetlands. There is a lack of uniform data collection and sparse research on correlations between satellite data and in situ measurements. Remote sensing (RS) has shown great promise in environmental monitoring. This project attempts to use satellite data and correlate metrics with in situ observations collected at five estuaries. Images for satellite data were processed to calculate 7 bands (SIs) using Python. Average SI values were calculated per month for 23 years. Publicly available data from 6 sites at ELK was used to obtain 10 parameters (OPs). Average OP values were calculated per month for 23 years. Linear correlations between the 7 SIs and 10 OPs were made and found to be inadequate (correlation = 1 to 64%). Fourier transform analysis on 7 SIs was performed. Dominant frequencies and amplitudes were extracted for 7 SIs, and a machine learning(ML) model was trained, validated, and tested for 10 OPs. Better correlations were observed between SIs and OPs, with certain time delays (0, 3, 4, 6 month delay), and ML was again performed. The OPs saw improved R² values in the range of 0.2 to 0.93. This approach can be used to get periodic analyses of overall wetland health with satellite indices. It proves that remote sensing can be used to develop correlations with critical parameters that measure eutrophication in situ data and can be used by practitioners to easily monitor wetland health.

Keywords: estuary, remote sensing, machine learning, Fourier transform

Procedia PDF Downloads 100

Authors: Laura M. F. Bertens

Abstract:

There is a long tradition of visually representing the art historical canon, in schematic overviews and diagrams. This is indicative of the desire for scientific, ‘objective’ knowledge of the kind (seemingly) produced in the natural sciences. These diagrams will, however, always retain an element of subjectivity and the modelling methods colour our perception of the represented information. In recent decades visualisations of art historical data, such as hand-drawn diagrams in textbooks, have been extended to include digital, computational tools. These tools significantly increase modelling strength and functionality. As such, they might be used to deconstruct and amend the very problem caused by traditional visualisations of the canon. In this paper, the use of digital tools for modelling the art historical canon is studied, in order to draw attention to the artificial nature of the static models that art historians are presented with in textbooks and lectures, as well as to explore the potential of digital, dynamic tools in creating new models. To study the way diagrams of the canon mediate the represented information, two modelling methods have been used on two case studies of existing diagrams. The tree diagram Stammbaum der neudeutschen Kunst (1823) by Ferdinand Olivier has been translated to a social network using the program Visone, and the famous flow chart Cubism and Abstract Art (1936) by Alfred Barr has been translated to an ontological model using Protégé Ontology Editor. The implications of the modelling decisions have been analysed in an art historical context. The aim of this project has been twofold. On the one hand the translation process makes explicit the design choices in the original diagrams, which reflect hidden assumptions about the Western canon. Ways of organizing data (for instance ordering art according to artist) have come to feel natural and neutral and implicit biases and the historically uneven distribution of power have resulted in underrepresentation of groups of artists. Over the last decades, scholars from fields such as Feminist Studies, Postcolonial Studies and Gender Studies have considered this problem and tried to remedy it. The translation presented here adds to this deconstruction by defamiliarizing the traditional models and analysing the process of reconstructing new models, step by step, taking into account theoretical critiques of the canon, such as the feminist perspective discussed by Griselda Pollock, amongst others. On the other hand, the project has served as a pilot study for the use of digital modelling tools in creating dynamic visualisations of the canon for education and museum purposes. Dynamic computer models introduce functionalities that allow new ways of ordering and visualising the artworks in the canon. As such, they could form a powerful tool in the training of new art historians, introducing a broader and more diverse view on the traditional canon. Although modelling will always imply a simplification and therefore a distortion of reality, new modelling techniques can help us get a better sense of the limitations of earlier models and can provide new perspectives on already established knowledge.

Keywords: canon, ontological modelling, Protege Ontology Editor, social network modelling, Visone

Procedia PDF Downloads 123

Authors: Prashansa Singh, Rohit Bajaj, Manjot Kaur

Abstract:

In the modern agricultural landscape, precision farming has emerged as a pivotal strategy for enhancing crop yield and optimizing resource utilization. This paper introduces an innovative Crop Recommendation System (CRS) that leverages the Internet of Things (IoT) technology and the Message Queuing Telemetry Transport (MQTT) protocol to collect critical environmental and soil data via sensors deployed across agricultural fields. The system is designed to address the challenges of real-time data acquisition, efficient data transmission, and dynamic crop recommendation through the application of advanced Artificial Intelligence (AI) and Machine Learning (ML) models. The CRS architecture encompasses a network of sensors that continuously monitor environmental parameters such as temperature, humidity, soil moisture, and nutrient levels. This sensor data is then transmitted to a central MQTT server, ensuring reliable and low-latency communication even in bandwidth-constrained scenarios typical of rural agricultural settings. Upon reaching the server, the data is processed and analyzed by AI/ML models trained to correlate specific environmental conditions with optimal crop choices and cultivation practices. These models consider historical crop performance data, current agricultural research, and real-time field conditions to generate tailored crop recommendations. This implementation gets 99% accuracy.

Keywords: Iot, MQTT protocol, machine learning, sensor, publish, subscriber, agriculture, humidity

Procedia PDF Downloads 62

Authors: Haq Nawaz, Muhammad Amjad Iqbal, Kamran Malik

Abstract:

"Altasreef" is an automated system available for Web and Android users which provide facility to the users to learn the Quran verbs. It provides the facility to the users to practice the learned material and also provide facility of exams of Arabic verbs variation focusing on Quran text. Arabic is a highly inflectional language. Almost all of its words connect to roots of three, four or five letters which approach the meaning of all their inflectional forms. In Arabic, a verb is formed by inserting the consonants into one of a set of verb patterns. Suffixes and prefixes are then added to generate the meaning of number, person, and gender. The active/passive voice and perfective aspect and other patterns are than generated. This application is designed for learners of Quranic Arabic who already have learn basics of Arabic conjugation. Application also provides the facility of translation of generated patterns. These translations are generated with the help of rule-based approach to give 100% results to the learners.

Keywords: NLP, Quran, Computational Linguistics, E Learning

Procedia PDF Downloads 164

Authors: Anudeep Appe, Bhanu Poluparthi, Lakshmi Kasivajjula, Udai Mv, Sobha Bagadi, Punya Modi, Aditya Singh, Hemanth Gunupudi, Spenser Troiano, Jeff Paul, Justin Stovall, Justin Yamamoto

Abstract:

The necessity of data-driven decisions in healthcare strategy formulation is rapidly increasing. A reliable framework which helps identify factors impacting a healthcare provider facility or a hospital (from here on termed as facility) market share is of key importance. This pilot study aims at developing a data-driven machine learning-regression framework which aids strategists in formulating key decisions to improve the facility’s market share which in turn impacts in improving the quality of healthcare services. The US (United States) healthcare business is chosen for the study, and the data spanning 60 key facilities in Washington State and about 3 years of historical data is considered. In the current analysis, market share is termed as the ratio of the facility’s encounters to the total encounters among the group of potential competitor facilities. The current study proposes a two-pronged approach of competitor identification and regression approach to evaluate and predict market share, respectively. Leveraged model agnostic technique, SHAP, to quantify the relative importance of features impacting the market share. Typical techniques in literature to quantify the degree of competitiveness among facilities use an empirical method to calculate a competitive factor to interpret the severity of competition. The proposed method identifies a pool of competitors, develops Directed Acyclic Graphs (DAGs) and feature level word vectors, and evaluates the key connected components at the facility level. This technique is robust since its data-driven, which minimizes the bias from empirical techniques. The DAGs factor in partial correlations at various segregations and key demographics of facilities along with a placeholder to factor in various business rules (for ex. quantifying the patient exchanges, provider references, and sister facilities). Identified are the multiple groups of competitors among facilities. Leveraging the competitors' identified developed and fine-tuned Random Forest Regression model to predict the market share. To identify key drivers of market share at an overall level, permutation feature importance of the attributes was calculated. For relative quantification of features at a facility level, incorporated SHAP (SHapley Additive exPlanations), a model agnostic explainer. This helped to identify and rank the attributes at each facility which impacts the market share. This approach proposes an amalgamation of the two popular and efficient modeling practices, viz., machine learning with graphs and tree-based regression techniques to reduce the bias. With these, we helped to drive strategic business decisions.

Keywords: competition, DAGs, facility, healthcare, machine learning, market share, random forest, SHAP

Procedia PDF Downloads 89