Search results for: user classification accuracy

Authors: Yang Yang, Dan Liu

Abstract:

Anomaly detection models for network flow based on machine learning have poor detection performance under extremely unbalanced training data conditions and also have slow detection speed and large resource consumption when deploying on network edge devices. Embedding multi-teacher knowledge distillation (mKD) in anomaly detection can transfer knowledge from multiple teacher models to a single model. Inspired by this, we proposed a state-of-the-art model, mKDNAD, to improve detection performance. mKDNAD mine and integrate the knowledge of one-dimensional sequence and two-dimensional image implicit in network flow to improve the detection accuracy of small sample classes. The multi-teacher knowledge distillation method guides the train of the student model, thus speeding up the model's detection speed and reducing the number of model parameters. Experiments in the CICIDS2017 dataset verify the improvements of our method in the detection speed and the detection accuracy in dealing with the small sample classes.

Keywords: network flow anomaly detection (NAD), multi-teacher knowledge distillation, machine learning, deep learning

Procedia PDF Downloads 116

Authors: Chao-Linag Hsieh, Zheng-Wei Ye, Chen-Kang Huang, Yeun-Chung Lee, Chih-Hong Sun, Tzai-Hung Wen, Jehn-Yih Juang, Joe-Air Jiang

Abstract:

Air pollution and climate warming become more and more intensified in many areas, especially in urban areas. Environmental parameters are critical information to air pollution and weather monitoring. Thus, it is necessary to develop a suitable air pollution and weather monitoring system for urban areas. In this study, a vehicle monitoring system (VMS) based on the IoT technique is developed. Cars are selected as the research tool because it can reach a greater number of streets to collect data. The VMS can monitor different environmental parameters, including ambient temperature and humidity, and air quality parameters, including PM2.5, NO₂, CO, and O₃. The VMS can provide other information, including GPS signals and the vibration information through driving a car on the street. Different sensor modules are used to measure the parameters and collect the measured data and transmit them to a cloud server through the LoRa protocol. A user interface is used to show the sensing data storing at the cloud server. To examine the performance of the system, a researcher drove a Nissan x-trail 1998 to the area close to the Da’an District office in Taipei to collect monitoring data. The collected data are instantly shown on the user interface. The four kinds of information are provided by the interface: GPS positions, weather parameters, vehicle information, and air quality information. With the VMS, users can obtain the information regarding air quality and weather conditions when they drive their car to an urban area. Also, government agencies can make decisions on traffic planning based on the information provided by the proposed VMS.

Keywords: LoRa, monitoring system, smart city, vehicle

Procedia PDF Downloads 408

Authors: Maximilian Elsen, Frank Tietze

Abstract:

The development and diffusion of green technologies are crucial for achieving our ambitious climate targets. The Paris Agreement commits its members to develop strategies for achieving net zero greenhouse gas emissions by the second half of the century. Governments, executives, and academics are working on net-zero strategies and the business of rating organisations on their environmental, social and governance (ESG) performance has grown tremendously in its public interest. ESG data is now commonly integrated into traditional investment analysis and an important factor in investment decisions. Creating these metrics, however, is inherently challenging as environmental and social impacts are hard to measure and uniform requirements on ESG reporting are lacking. ESG metrics are often incomplete and inconsistent as they lack fully accepted reporting standards and are often of qualitative nature. This study explores the use of patent data for assessing the environmental performance of companies by focusing on their patented inventions in the space of climate change mitigation and adaptation technologies (CCMAT). The present study builds on the successful identification of CCMAT patents. In this context, the study adopts the Y02 patent classification, a fully cross-sectional tagging scheme that is fully incorporated in the Cooperative Patent Classification (CPC), to identify Climate Change Adaptation Technologies. The Y02 classification was jointly developed by the European Patent Office (EPO) and the United States Patent and Trademark Office (USPTO) and provides means to examine technologies in the field of mitigation and adaptation to climate change across relevant technologies. This paper develops sustainability-related metrics for firm-level patent portfolios. We do so by adopting a three-step approach. First, we identify relevant CCMAT patents based on their classification as Y02 CPC patents. Second, we examine the technological strength of the identified CCMAT patents by including more traditional metrics from the field of patent analytics while considering their relevance in the space of CCMAT. Such metrics include, among others, the number of forward citations a patent receives, as well as the backward citations and the size of the focal patent family. Third, we conduct our analysis on a firm level by sector for a sample of companies from different industries and compare the derived sustainability performance metrics with the firms’ environmental and financial performance based on carbon emissions and revenue data. The main outcome of this research is the development of sustainability-related metrics for firm-level environmental performance based on patent data. This research has the potential to complement existing ESG metrics from an innovation perspective by focusing on the environmental performance of companies and putting them into perspective to conventional financial performance metrics. We further provide insights into the environmental performance of companies on a sector level. This study has implications of both academic and practical nature. Academically, it contributes to the research on eco-innovation and the literature on innovation and intellectual property (IP). Practically, the study has implications for policymakers by deriving meaningful insights into the environmental performance from an innovation and IP perspective. Such metrics are further relevant for investors and potentially complement existing ESG data.

Keywords: climate change mitigation, innovation, patent portfolios, sustainability

Procedia PDF Downloads 81

Authors: Shibei Huang

Abstract:

In deep energy retrofits, the thermal performance of existing building envelopes is often difficult to determine with a high level of accuracy. For older buildings, the records of existing assemblies are often incomplete or inaccurate. To obtain greater baseline performance accuracy for energy models, in-field measurement tools can be used to obtain data on the thermal performance of the existing assemblies. For a known assembly, these field measurements assist in validating the U-factor estimates. If the field-measured U-factor consistently varies from the calculated prediction, those measurements prompt further study. For an unknown assembly, successful field measurements can provide approximate U-factor evaluation, validate assumptions, or identify anomalies requiring further investigation. Using case studies, this presentation will focus on the non-destructive methods utilizing a set of various field tools to validate the baseline U-factors for a range of existing buildings with various wall assemblies. The lessons learned cover what can be achieved, the limitations of these approaches and tools, and ideas for improving the validity of measurements. Key factors include the weather conditions, the interior conditions, the thermal mass of the measured assemblies, and the thermal profiles of the assemblies in question.

Keywords: existing building, sensor, thermal analysis, retrofit

Procedia PDF Downloads 59

Authors: Abdul Basit Kiani

Abstract:

Web development, the art of creating and maintaining websites, has witnessed remarkable advancements. The aim is to provide an overview of some of the cutting-edge developments in the field. Firstly, the rise of responsive web design has revolutionized user experiences across devices. With the increasing prevalence of smartphones and tablets, web developers have adapted to ensure seamless browsing experiences, regardless of screen size. This progress has greatly enhanced accessibility and usability, catering to the diverse needs of users worldwide. Additionally, the evolution of web frameworks and libraries has significantly streamlined the development process. Tools such as React, Angular, and Vue.js have empowered developers to build dynamic and interactive web applications with ease. These frameworks not only enhance efficiency but also bolster scalability, allowing for the creation of complex and feature-rich web solutions. Furthermore, the emergence of progressive web applications (PWAs) has bridged the gap between native mobile apps and web development. PWAs leverage modern web technologies to deliver app-like experiences, including offline functionality, push notifications, and seamless installation. This innovation has transformed the way users interact with websites, blurring the boundaries between traditional web and mobile applications. Moreover, the integration of artificial intelligence (AI) and machine learning (ML) has opened new horizons in web development. Chatbots, intelligent recommendation systems, and personalization algorithms have become integral components of modern websites. These AI-powered features enhance user engagement, provide personalized experiences, and streamline customer support processes, revolutionizing the way businesses interact with their audiences. Lastly, the emphasis on web security and privacy has been a pivotal area of progress. With the increasing incidents of cyber threats, web developers have implemented robust security measures to safeguard user data and ensure secure transactions. Innovations such as HTTPS protocol, two-factor authentication, and advanced encryption techniques have bolstered the overall security of web applications, fostering trust and confidence among users. Hence, recent progress in web development has propelled the industry forward, enabling developers to craft innovative and immersive digital experiences. From responsive design to AI integration and enhanced security, the landscape of web development continues to evolve, promising a future filled with endless possibilities.

Keywords: progressive web applications (PWAs), web security, machine learning (ML), web frameworks, advancement responsive web design

Procedia PDF Downloads 51

Authors: Xiang Shijie, Zhou Dong, Tian Dan

Abstract:

Speaker recognition is a crucial task in the field of speech processing, aimed at identifying individuals based on their vocal characteristics. However, existing speaker recognition methods face numerous challenges. Traditional methods primarily rely on audio signals, which often suffer from limitations in noisy environments, variations in speaking style, and insufficient sample sizes. Additionally, relying solely on audio features can sometimes fail to capture the unique identity of the speaker comprehensively, impacting recognition accuracy. To address these issues, we propose a multi-modal network architecture that simultaneously processes both audio and text signals. By gradually integrating audio and text features, we leverage the strengths of both modalities to enhance the robustness and accuracy of speaker recognition. Our experiments demonstrate significant improvements with this multi-modal approach, particularly in complex environments, where recognition performance has been notably enhanced. Our research not only highlights the limitations of current speaker recognition methods but also showcases the effectiveness of multi-modal fusion techniques in overcoming these limitations, providing valuable insights for future research.

Keywords: feature fusion, memory network, multimodal input, speaker recognition

Procedia PDF Downloads 18

Authors: Néstor Iván Rojas, Wilson Medina Sierra

Abstract:

Managing spatial information is described through a Geographic Information System (GIS), for some neighborhoods in the city of Tunja, in relation to the structural typology of the buildings. The use of GIS provides tools that facilitate the capture, processing, analysis and dissemination of cartographic information, product quality evaluation of the classification of buildings. Allows the development of a method that unifies and standardizes processes information. The project aims to generate a geographic database that is useful to the entities responsible for planning and disaster prevention and care for vulnerable populations, also seeks to be a basis for seismic vulnerability studies that can contribute in a study of urban seismic microzonation. The methodology consists in capturing the plat including road naming, neighborhoods, blocks and buildings, to which were added as attributes, the product of the evaluation of each of the housing data such as the number of inhabitants and classification, year of construction, the predominant structural systems, the type of mezzanine board and state of favorability, the presence of geo-technical problems, the type of cover, the use of each building, damage to structural and non-structural elements . The above data are tabulated in a spreadsheet that includes cadastral number, through which are systematically included in the respective building that also has that attribute. Geo-referenced data base is obtained, from which graphical outputs are generated, producing thematic maps for each evaluated data, which clearly show the spatial distribution of the information obtained. Using GIS offers important advantages for spatial information management and facilitates consultation and update. Usefulness of the project is recognized as a basis for studies on issues of planning and prevention.

Keywords: microzonation, buildings, geo-processing, cadastral number

Procedia PDF Downloads 330

Authors: Farbod Behnaminia, Saeed Samet

Abstract:

The COVID-19 pandemic has paralyzed many lives until a vaccine was available, which caused the so-called “new normal.” According to the World Health Organization (WHO), COVID-19 is an infectious disease. It can cause significant illness or death in anyone. Governments and health officials tried to impose rules and regulations to avoid and slow down transmission. Therefore, software engineers worldwide developed applications to trace and track patients’ movements and notify others, mainly using Bluetooth. In this way, everyone could be informed whether they come in close contact with someone who has COVID-19 and takes proper safety precautions. Because most of the applications use technologies that can potentially reveal the user’s identity and location, researchers have debated privacy preservation and how to improve user privacy during such pandemics. Thanks to Distributed Ledger Technology (DLT), there have been some proposed methods to develop privacy-preserving Patient Tracking Systems in the last two years. As an instance of the DLT, Blockchain is like a decentralized peer-to-peer database that maintains a record of transactions. Transactions are immutable, transparent, and anonymous in this system. We conducted a comprehensive evaluation of the literature by looking for papers in the relevant field and dividing them into pre- and post-pandemic systems. Additionally, we discussed the many uses of blockchain technology in pandemic control. We found that two major obstacles facing blockchain implementation across many healthcare systems are scalability and privacy. The Polkadot platform is presented, along with a review of its efficacy in tackling current concerns. A more scalable healthcare system is achievable in the near future using Polkadot as well as a much more privacy-preserving environment.

Keywords: blockchain, electronic record management, EHR, privacy-preserving, patient tracking, COVID-19, trust and confidence, Polkadot

Procedia PDF Downloads 99

Authors: Chuks Odiegwu-Enwerem

Abstract:

The classical notion of opinion leadership presupposes that the media is at the center of an effective and successful opinion leadership. Under this idea, an opinion leader is an active media user who consumes, understands, digests and interprets the messages for the understanding and acceptance/adoption by lower-end media users – whose access and understanding of media content are supposedly low. Because of their unique access to and presumed understanding of media functions and their content, opinion leaders are typically esteemed by those who look forward to and accept their opinions. Lazarsfeld and Katz’s two-step flow of communication theory is the basis of opinion leadership – propelled by limited access to the media. With the emergence and spread of social media and its unlimited access by all and sundry, however, the study interrogates the relevance and application of opinion leaders and, by implication, the two-step flow communication theory in Nigeria’s Northcentral region. It seeks to determine whether opinion leaders still exist in the picture and if they still exert considerable influence, especially in matters of political conversations and decision-making among the citizens of this area. It further explores whether the diffusion of social media is a reality and how the ‘low-end’ media users react to the new-found freedom of access to media, and how they are using it to inform their decisions on important matters as well as examines if they are still glued to their opinion leaders. This study explores the empirical dimensions of the two-step flow hypothesis in relation to the activities of social media to determine if a change has occurred and in what direction, using mixed methos of Survey and in-depth interviews. Our understanding and belief in some theoretical assumptions may be enhanced or challenged by the study outcome.

Keywords: Opinion Leadership, Active Media User, Two-Step-Flow, Social media, Northcentral Nigeria

Procedia PDF Downloads 68

Authors: Apoorva Vinod, Archana Mathur, Snehanshu Saha

Abstract:

Though there exists a plethora of Activation Functions (AFs) used in single and multiple hidden layer Neural Networks (NN), their behavior always raised curiosity, whether used in combination or singly. The popular AFs –Sigmoid, ReLU, and Tanh–have performed prominently well for shallow and deep architectures. Most of the time, AFs are used singly in multi-layered NN, and, to the best of our knowledge, their performance is never studied and analyzed deeply when used in combination. In this manuscript, we experiment with multi-layered NN architecture (both on shallow and deep architectures; Convolutional NN and VGG16) and investigate how well the network responds to using two different AFs (Sigmoid-Tanh, Tanh-ReLU, ReLU-Sigmoid) used alternately against a traditional, single (Sigmoid-Sigmoid, Tanh-Tanh, ReLUReLU) combination. Our results show that using two different AFs, the network achieves better accuracy, substantially lower loss, and faster convergence on 4 computer vision (CV) and 15 Non-CV (NCV) datasets. When using different AFs, not only was the accuracy greater by 6-7%, but we also accomplished convergence twice as fast. We present a case study to investigate the probability of networks suffering vanishing and exploding gradients when using two different AFs. Additionally, we theoretically showed that a composition of two or more AFs satisfies Universal Approximation Theorem (UAT).

Keywords: activation function, universal approximation function, neural networks, convergence

Procedia PDF Downloads 153

Authors: Shreya Borthakur, Aastha Vartak

Abstract:

This Electroencephalogram (EEG) project on Rapid Visual Serial Processing (RSVP) paradigm explores the temporal dynamics of visual attention processing in response to rapidly presented visual stimuli. The study builds upon previous research that used real-world images in RSVP tasks to understand the emergence of object representations in the human brain. The objectives of the research include investigating the differences in accuracy and reaction times between 5 Hz and 20 Hz presentation rates, as well as examining the prominent brain waves, particularly alpha and beta waves, associated with the attention task. The pre-processing and data analysis involves filtering EEG data, creating epochs for target stimuli, and conducting statistical tests using MATLAB, EEGLAB, Chronux toolboxes, and R. The results support the hypotheses, revealing higher accuracy at a slower presentation rate, faster reaction times for less complex targets, and the involvement of alpha and beta waves in attention and cognitive processing. This research sheds light on how short-term memory and cognitive control affect visual processing and could have practical implications in fields like education.

Keywords: RSVP, attention, visual processing, attentional blink, EEG

Procedia PDF Downloads 66

Authors: Samad Sajjadi

Abstract:

Machine translation is a new subject increasingly being used by academic writers, especially students and researchers whose native language is not English. There are numerous studies conducted on machine translation, but few investigations have assessed the accuracy of machine translation from Persian to English at lexical, semantic, and syntactic levels. Using Groves and Mundt’s (2015) Model of error taxonomy, the current study evaluated Persian-to-English translations produced by two famous online translators, Google Translate and Bing Translator. A total of 240 texts were randomly selected from different academic fields (law, literature, medicine, and mass media), and 60 texts were considered for each domain. All texts were rendered by the two translation systems and then by four human translators. All statistical analyses were applied using SPSS. The results indicated that Google translations were more accurate than the translations produced by the Bing Translator, especially in the domains of medicine (lexis: 186 vs. 225; semantic: 44 vs. 48; syntactic: 148 vs. 264 errors) and mass media (lexis: 118 vs. 149; semantic: 25 vs. 32; syntactic: 110 vs. 220 errors), respectively. Nonetheless, both machines are reasonably accurate in Persian-to-English translation of lexicons and syntactic structures, particularly from mass media and medical texts.

Keywords: machine translations, accuracy, human translation, efficiency

Procedia PDF Downloads 73

Authors: Sangguk Ahn, Youngyu Kim, Jaheon Gu, Gyoutae Park

Abstract:

In this paper, we developed a dual gas leak instrument to detect Hydrocarbon (HC) and Monoxide (CO) gases. To two kinds of gases, it is necessary to design compact structure for sensors. And then it is important to draw sensing circuits such as measuring, amplifying and filtering. After that, it should be well programmed with robust, systematic and module coding methods. In center of them, improvement of accuracy and initial response time are a matter of vital importance. To manufacture distinguished gas leak detector, we applied intrinsically safe explosion-proof structure to lithium ion battery, main circuits, a pump with motor, color LCD interfaces and sensing circuits. On software, to enhance measuring accuracy we used numerical analysis such as Lagrange and Neville interpolation. Performance test result is conducted by using standard Methane with seven different concentrations with three other products. We want raise risk prevention and efficiency of gas safe management through distributing to the field of gas safety. Acknowledgment: This study was supported by Small and Medium Business Administration under the research theme of ‘Commercialized Development of a portable intrinsically safe explosion-proof type dual gas leak detector’, (task number S2456036).

Keywords: gas leak, dual gas detector, intrinsically safe, explosion proof

Procedia PDF Downloads 226

Authors: Luiz Daniel Garay Trindade, Antonio De Freitas Valle Neto, Fabio Paulo Basso, Elder De Macedo Rodrigues, Maicon Bernardino, Daniel Welfer, Daniel Muller

Abstract:

Agriculture plays an important role in society since it is one of the main sources of food in the world. To help the production and yield of crops, precision agriculture makes use of technologies aiming at improving productivity and quality of agricultural commodities. One of the problems hampering quality of agricultural production is the disease affecting crops. Failure in detecting diseases in a short period of time can result in small or big damages to production, causing financial losses to farmers. In order to provide a map of the contributions destined to the early detection of plant diseases and a comparison of the accuracy of the selected studies, a systematic literature review of the literature was performed, showing techniques for digital image processing and neural networks. We found 35 interesting tool support alternatives to detect disease in 19 plants. Our comparison of these studies resulted in an overall average accuracy of 87.45%, with two studies very closer to obtain 100%.

Keywords: pattern recognition, image processing, deep learning, precision agriculture, smart farming, agricultural automation

Procedia PDF Downloads 374

Authors: Faith-Michael E. Uzoka, Boluwaji A. Akinnuwesi, Taiwo Amoo, Flora Aladi, Stephen Fashoto, Moses Olaniyan, Joseph Osuji

Abstract:

The overarching aim of this study is to develop a soft-computing system for the differential diagnosis of tropical diseases. These conditions are of concern to health bodies, physicians, and the community at large because of their mortality rates, and difficulties in early diagnosis due to the fact that they present with symptoms that overlap, and thus become ‘confusable’. We report on the first phase of our study, which focuses on the development of a fuzzy cognitive map model for early differential diagnosis of tropical diseases. We used malaria as a case disease to show the effectiveness of the FCM technology as an aid to the medical practitioner in the diagnosis of tropical diseases. Our model takes cognizance of manifested symptoms and other non-clinical factors that could contribute to symptoms manifestations. Our model showed 85% accuracy in diagnosis, as against the physicians’ initial hypothesis, which stood at 55% accuracy. It is expected that the next stage of our study will provide a multi-disease, multi-symptom model that also improves efficiency by utilizing a decision support filter that works on an algorithm, which mimics the physician’s diagnosis process.

Keywords: medical diagnosis, tropical diseases, fuzzy cognitive map, decision support filters, malaria differential diagnosis

Procedia PDF Downloads 316

Authors: Usman Ahmed Khan

Abstract:

Urbanization directly affects the existing infrastructure, landscape modification, environmental contamination, and traffic pollution, especially if there is a lack of urban planning. Recently, the rapid urban sprawl has resulted in less developed green areas and has devastating environmental consequences. This study was aimed to study the past urban expansion rates and measure LST from satellite data. The land use land cover (LULC) maps of years 1996, 2010, 2013, and 2017 were generated using landsat satellite images. Four main classes, i.e., water, urban, bare land, and vegetation, were identified using unsupervised classification with iterative self-organizing data analysis (isodata) technique. The LST from satellite thermal data can be derived from different procedures: atmospheric, radiometric calibrations and surface emissivity corrections, classification of spatial changeability in land-cover. Different methods and formulas were used in the algorithm that successfully retrieves the land surface temperature to help us study the thermal environment of the ground surface. To verify the algorithm, the land surface temperature and the near-air temperature were compared. The results showed that, From 1996-2017, urban areas increased to about a considerable increase of about 48%. Few areas of the city also shown in a reduction in LST from the year 1996-2017 that actually began their transitional phase from rural to urban LULC. The mean temperature of the city increased averagely about 1ºC each year in the month of October. The green and vegetative areas witnessed a decrease in the area while a higher number of pixels increased in urban class.

Keywords: LST, LULC, isodata, urbanization

Procedia PDF Downloads 99

Authors: David Sinfield, Thomas Cochrane, Marcos Steagall

Abstract:

While there is much hype around the potential and development of mobile virtual reality (VR), the two key critical success factors are the ease of user experience and the development of a simple user-generated content ecosystem. Educational technology history is littered with the debris of over-hyped revolutionary new technologies that failed to gain mainstream adoption or were quickly superseded. Examples include 3D television, interactive CDROMs, Second Life, and Google Glasses. However, we argue that this is the result of curriculum design that substitutes new technologies into pre-existing pedagogical strategies that are focused upon teacher-delivered content rather than exploring new pedagogical strategies that enable student-determined learning or heutagogy. Visual Communication design based learning such as Graphic Design, Illustration, Photography and Design process is heavily based on the traditional forms of the classroom environment whereby student interaction takes place both at peer level and indeed teacher based feedback. In doing so, this makes for a healthy creative learning environment, but does raise other issue in terms of student to teacher learning ratios and reduced contact time. Such issues arise when students are away from the classroom and cannot interact with their peers and teachers and thus we see a decline in creative work from the student. Using AR and VR as a means of stimulating the students and to think beyond the limitation of the studio based classroom this paper will discuss the outcomes of a student project considering the virtual classroom and the techniques involved. The Atelier learning environment is especially suited to the Visual Communication model as it deals with the creative processing of ideas that needs to be shared in a collaborative manner. This has proven to have been a successful model over the years, in the traditional form of design education, but has more recently seen a shift in thinking as we move into a more digital model of learning and indeed away from the classical classroom structure. This study focuses on the outcomes of a student design project that employed Augmented Reality and Virtual Reality technologies in order to expand the dimensions of the classroom beyond its physical limits. Augmented Reality when integrated into the learning experience can improve the learning motivation and engagement of students. This paper will outline some of the processes used and the findings from the semester-long project that took place.

Keywords: augmented reality, blogging, design in community, enhanced learning and teaching, graphic design, new technologies, virtual reality, visual communications

Procedia PDF Downloads 237

Authors: Ayliana Dharmawan, Heng Yong Sheng, Zhang Xiaojin, Tan Thai Lian

Abstract:

To facilitate discharge planning, doctors are presently required to assign an Estimated Discharge Date (EDD) for each patient admitted to the hospital. This assignment of the EDD is largely based on the doctor’s judgment. This can be difficult for cases which are complex or relatively new to the doctor. It is hypothesized that a data-driven approach would be able to facilitate the doctors to make accurate estimations of the discharge date. Making use of routinely collected data on inpatient discharges between January 2013 and May 2016, a predictive model was developed using machine learning techniques to predict the Length of Stay (and hence the EDD) of inpatients, at the point of admission. The predictive performance of the model was compared to that of the clinicians using accuracy measures. Overall, the best performing model was found to be able to predict EDD with an accuracy improvement in Average Squared Error (ASE) by -38% as compared to the first EDD determined by the present method. It was found that important predictors of the EDD include the provisional diagnosis code, patient’s age, attending doctor at admission, medical specialty at admission, accommodation type, and the mean length of stay of the patient in the past year. The predictive model can be used as a tool to accurately predict the EDD.

Keywords: inpatient, estimated discharge date, EDD, prediction, data-driven

Procedia PDF Downloads 170

Authors: Marden Silva, Danielle Guerra

Abstract:

Some studies have indicated that pronunciation teaching hasn’t been paid enough attention by teachers regarding EFL contexts. In particular, segmental and suprasegmental features through genre-based approach may be an opportunity on how to integrate pronunciation into a more meaningful learning practice. Therefore, the aim of this project was to carry out a survey on some aspects related to English pronunciation that Brazilian students consider more difficult to learn, thus enabling the discussion of strategies that can facilitate the development of oral skills in English classes by integrating the teaching of phonetic-phonological aspects into the genre-based approach. Notions of intelligibility, fluency and accuracy were proposed by some authors as an ideal didactic sequence. According to their proposals, basic learners should be exposed to activities focused on the notion of intelligibility as well as intermediate students to the notion of fluency, and finally more advanced ones to accuracy practices. In order to test this hypothesis, data collection was conducted during three high school English classes at Federal Center for Technological Education of Minas Gerais (CEFET-MG), in Brazil, through questionnaires and didactic activities, which were recorded and transcribed for further analysis. The genre debate was chosen to facilitate the oral expression of the participants in a freer way, making them answering questions and giving their opinion about a previously selected topic. The findings indicated that basic students demonstrated more difficulty with aspects of English pronunciation than the others. Many of the intelligibility aspects analyzed had to be listened more than once for a better understanding. For intermediate students, the speeches recorded were considerably easier to understand, but nevertheless they found it more difficult to pronounce the words fluently, often interrupting their speech to think about what they were going to say and how they would talk. Lastly, more advanced learners seemed to express their ideas more fluently, but still subtle errors related to accuracy were perceptible in speech, thereby confirming the proposed hypothesis. It was also seen that using genre-based approach to promote oral communication in English classes might be a relevant method, considering the socio-communicative function inherent in the suggested approach.

Keywords: EFL, genre-based approach, oral skills, pronunciation

Procedia PDF Downloads 126

Authors: Michael Smalenberger

Abstract:

Intelligent tutoring systems (ITS) are computer-based platforms which can incorporate artificial intelligence to provide step-by-step guidance as students practice problem-solving skills. ITS can replicate and even exceed some benefits of one-on-one tutoring, foster transactivity in collaborative environments, and lead to substantial learning gains when used to supplement the instruction of a teacher or when used as the sole method of instruction. A common facet of many ITS is their use of Bayesian Knowledge Tracing (BKT) to estimate parameters necessary for the implementation of the artificial intelligence component, and for the probability of mastery of a knowledge component relevant to the ITS. While various techniques exist to estimate these parameters and probability of mastery, none directly and reliably ask the user to self-assess these. In this study, 111 undergraduate students used an ITS in a college-level introductory statistics course for which detailed transaction-level observations were recorded, and users were also routinely asked direct questions that would lead to such a self-assessment. Comparisons were made between these self-assessed values and those obtained using commonly used estimation techniques. Our findings show that such self-assessments are particularly relevant at the early stages of ITS usage while transaction level data are scant. Once a user’s transaction level data become available after sufficient ITS usage, these can replace the self-assessments in order to eliminate the identifiability problem in BKT. We discuss how these findings are relevant to the number of exercises necessary to lead to mastery of a knowledge component, the associated implications on learning curves, and its relevance to instruction time.

Keywords: Bayesian Knowledge Tracing, Intelligent Tutoring System, in vivo study, parameter estimation

Procedia PDF Downloads 170

Authors: Amirhossein Mohajerzadeh, Abolghasem Mohajerzadeh

Abstract:

In this paper, target parameter is estimated with desirable precision in hierarchical wireless sensor networks (WSN) while the proposed algorithm also tries to prolong network lifetime as much as possible, using efficient data collecting algorithm. Target parameter distribution function is considered unknown. Sensor nodes sense the environment and send the data to the base station called fusion center (FC) using hierarchical data collecting algorithm. FC builds underlying phenomena based on collected data. Considering the aggregation level, x, the goal is providing the essential infrastructure to find the best value for aggregation level in order to prolong network lifetime as much as possible, while desirable accuracy is guaranteed (required sample size is fully depended on desirable precision). First, the sample size calculation algorithm is discussed, second, the average queue length based on M/M[x]/1/K queue model is determined and it is used for energy consumption calculation. Nodes can decrease transmission cost by aggregating incoming data. Furthermore, the performance of the new algorithm is evaluated in terms of lifetime and estimation accuracy.

Keywords: aggregation, estimation, queuing, wireless sensor network

Procedia PDF Downloads 183

Authors: Ibrahim A. Adeniran, Rui F. Zhu, Man S. Wong

Abstract:

Owing to the insufficiency in the spatial representativeness and continuity of in situ temperature measurements from weather stations (WS), the use of temperature measurement from WS for large-range diurnal analysis in heterogenous landscapes has been limited. This has made the accurate estimation of land surface temperature (LST) from remotely sensed data more crucial. Moreover, the study of dynamic interaction between the atmosphere and the physical surface of the Earth could be enhanced at both annual and diurnal scales by using optimal LST data derived from satellite sensors. The tradeoff between the spatial and temporal resolution of LSTs from satellite’s thermal infrared sensors (TIRS) has, however, been a major challenge, especially when high spatiotemporal LST data are recommended. It is well-known from existing literature that polar satellites have the advantage of high spatial resolution, while geostationary satellites have a high temporal resolution. Hence, this study is aimed at designing a framework for the cross-comparison of LST data from polar and geostationary satellites in a heterogeneous landscape. This could help to understand the relationship between the LST estimates from the two satellites and, consequently, their integration in diurnal LST analysis. Landsat-8 satellite data will be used as the representative of the polar satellite due to the availability of its long-term series, while the Himawari-8 satellite will be used as the data source for the geostationary satellite because of its improved TIRS. For the study area, Hong Kong Special Administrative Region (HK SAR) will be selected; this is due to the heterogeneity in the landscape of the region. LST data will be retrieved from both satellites using the Split window algorithm (SWA), and the resulting data will be validated by comparing satellite-derived LST data with temperature data from automatic WS in HK SAR. The LST data from the satellite data will then be separated based on the land use classification in HK SAR using the Global Land Cover by National Mapping Organization version3 (GLCNMO 2013) data. The relationship between LST data from Landsat-8 and Himawari-8 will then be investigated based on the land-use class and over different seasons of the year in order to account for seasonal variation in their relationship. The resulting relationship will be spatially and statistically analyzed and graphically visualized for detailed interpretation. Findings from this study will reveal the relationship between the two satellite data based on the land use classification within the study area and the seasons of the year. While the information provided by this study will help in the optimal combination of LST data from Polar (Landsat-8) and geostationary (Himawari-8) satellites, it will also serve as a roadmap in the annual and diurnal urban heat (UHI) analysis in Hong Kong SAR.

Keywords: automatic weather station, Himawari-8, Landsat-8, land surface temperature, land use classification, split window algorithm, urban heat island

Procedia PDF Downloads 71

Authors: Abdul Hakim Chikho

Abstract:

Calculating the ultimate load of a column in a sway framed structure involves, in the currently used design method, the calculation of the column effective length and utilizing the interaction formulas or tables. Therefore, no allowance is usually made for the effects of the presence of semi rigid connections or the presence of infill panels. In this paper, a new and simple design procedure is recommend to calculate the ultimate load of a framed Column allowing for the presence of rotational end restraints, semi rigid connections, the column end moments resulted from the applied vertical and horizontal loading and infill panels in real steel structure. In order to verify the accuracy of the recommended method to predict good and safe estimations of framed column ultimate loads, several examples have been solved utilizing the recommended procedure, and the results were compared to those obtained using a second order computer program, and good correlation had been obtained. Therefore, the accuracy of the proposed method to predict the Behaviour of practical steel columns in framed structures has been verified.

Keywords: column ultimate load, semi rigid connections, steel column, infill panel, steel structure

Procedia PDF Downloads 176

Authors: Ian Omung'a

Abstract:

Breast cancer remains to be one of the deadliest cancers for women worldwide, with the risk of developing tumors being as high as 50 percent in Sub-Saharan African countries like Kenya. With as many as 42 percent of these cases set to be diagnosed late when cancer has metastasized and or the prognosis has become terminal, Full Field Digital [FFD] Mammography remains an effective screening technique that leads to early detection where in most cases, successful interventions can be made to control or eliminate the tumors altogether. FFD Mammograms have been proven to multiply more effective when used together with Computer-Aided Detection and Diagnosis [CADe] systems, relying on algorithmic implementations of Deep Learning techniques in Computer Vision to carry out deep pattern recognition that is comparable to the level of a human radiologist and decipher whether specific areas of interest in the mammogram scan image portray abnormalities if any and whether these abnormalities are indicative of a benign or malignant tumor. Within this paper, we review emergent Deep Learning techniques that will prove relevant to the development of State-of-The-Art FFD Mammogram CADe systems. These techniques will span self-supervised learning for context-encoded occlusion, self-supervised learning for pre-processing and labeling automation, as well as the creation of a standardized large-scale mammography dataset as a benchmark for CADe systems' evaluation. Finally, comparisons are drawn between existing practices that pre-date these techniques and how the development of CADe systems that incorporate them will be different.

Keywords: breast cancer diagnosis, computer aided detection and diagnosis, deep learning, whole mammogram classfication, ultrasound classification, computer vision

Procedia PDF Downloads 91

Authors: Cheng Zhang, James Marco, Walid Allafi, Truong Q. Dinh, W. D. Widanage

Abstract:

Equivalent circuit models (ECMs) are widely used in battery management systems in electric vehicles and other battery energy storage systems. The battery dynamics and the model parameters vary under different working conditions, such as different temperature and state of charge (SOC) levels, and therefore online parameter identification can improve the modelling accuracy. This paper presents a way of online ECM parameter identification using a continuous time (CT) estimation method. The CT estimation method has several advantages over discrete time (DT) estimation methods for ECM parameter identification due to the widely separated battery dynamic modes and fast sampling. The presented method can be used for online SOC estimation. Test data are collected using a lithium ion cell, and the experimental results show that the presented CT method achieves better modelling accuracy compared with the conventional DT recursive least square method. The effectiveness of the presented method for online SOC estimation is also verified on test data.

Keywords: electric circuit model, continuous time domain estimation, linear integral filter method, parameter and SOC estimation, recursive least square

Procedia PDF Downloads 379

Authors: Marija Zulić, Vanda Hájková, Nina Brkić–Jovanović, Srećko Potić, Sanja Tomić

Abstract:

The name cerebral palsy comes from the word cerebrum, which means the brain and the word palsy, which means seizure, and essentially refers to the movement disorder. In the clinical picture of cerebral palsy, basic neuromotor disorders are associated with other various disorders: behavioural, intellectual, speech, sensory, epileptic seizures, and bone and joint deformities. Motor speech disorders are among the most common difficulties present in people with cerebral palsy. Social participation represents an interaction between an individual and their social environment. Quality of social participation of the students with cerebral palsy at school is an important indicator of their successful participation in adulthood. One of the most important skills for the undisturbed social participation is ability of good communication. The aim of the study was to determine relation between social participation of students with cerebral palsy and presence of their speech impairment in primary schools in the Czech Republic. The study was performed in the Czech Republic in mainstream schools and schools established for the pupils with special education needs. We analysed 75 children with cerebral palsy aged between six and twelve years attending up to sixth grade by using the first and the third part of the school function assessment questionnaire as the main instrument. The other instrument we used in the research is the Gross motor function classification system–five–level classification system, which measures degree of motor functions of children and youth with cerebral palsy. Funding for this study was provided by the Grant Agency of Charles University in Prague.

Keywords: cerebral palsy, social participation, speech disorders, The Czech Republic, the school function assessment

Procedia PDF Downloads 282

Authors: Jian Shi, Baoguo Yu, Haonan Jia, Meng Liu, Ping Huang

Abstract:

Adapting to the flexibility of war, a large number of guided weapons launch from aircraft. Therefore, the inertial navigation system loaded in the weapon needs to undergo an alignment process in the air. This article proposes the following methods to the problem of inaccurate modeling of the system under large misalignment angles, the accuracy reduction of filtering caused by outliers, and the noise changes in GPS signals: first, considering the large misalignment errors of Strapdown Inertial Navigation System (SINS)/GPS, a more accurate model is made rather than to make a small-angle approximation, and the Unscented Kalman Filter (UKF) algorithms are used to estimate the state; then, taking into account the impact of GPS noise changes on the fine alignment algorithm, the innovation adaptive filtering algorithm is introduced to estimate the GPS’s noise in real-time; at the same time, in order to improve the anti-interference ability of the air fine alignment algorithm, a robust filtering algorithm based on outlier detection is combined with the air fine alignment algorithm to improve the robustness of the algorithm. The algorithm can improve the alignment accuracy and robustness under interference conditions, which is verified by simulation.

Keywords: air alignment, fine alignment, inertial navigation system, integrated navigation system, UKF

Procedia PDF Downloads 161

Authors: Akeel A. Shah, Tong Zhang

Abstract:

Computational approaches to learning the properties of materials are commonplace, motivated by the need to screen or design materials for a given application, e.g., semiconductors and energy storage. Experimental approaches can be both time consuming and costly. Unfortunately, computational approaches such as ab-initio electronic structure calculations and classical or ab-initio molecular dynamics are themselves can be too slow for the rapid evaluation of materials, often involving thousands to hundreds of thousands of candidates. Machine learning assisted approaches have been developed to overcome the time limitations of purely physics-based approaches. These approaches, on the other hand, require large volumes of data for training (hundreds of thousands on many standard data sets such as QM7b). This means that they are limited by how quickly such a large data set of physics-based simulations can be established. At high fidelity, such as configuration interaction, composite methods such as G4, and coupled cluster theory, gathering such a large data set can become infeasible, which can compromise the accuracy of the predictions - many applications require high accuracy, for example band structures and energy levels in semiconductor materials and the energetics of charge transfer in energy storage materials. In order to circumvent this problem, multi-fidelity approaches can be adopted, for example the Δ-ML method, which learns a high-fidelity output from a low-fidelity result such as Hartree-Fock or density functional theory (DFT). The general strategy is to learn a map between the low and high fidelity outputs, so that the high-fidelity output is obtained a simple sum of the physics-based low-fidelity and correction, Although this requires a low-fidelity calculation, it typically requires far fewer high-fidelity results to learn the correction map, and furthermore, the low-fidelity result, such as Hartree-Fock or semi-empirical ZINDO, is typically quick to obtain, For high-fidelity outputs the result can be an order of magnitude or more in speed up. In this work, a new multi-fidelity approach is developed, based on a graph convolutional network (GCN) combined with semi-supervised learning. The GCN allows for the material or molecule to be represented as a graph, which is known to improve accuracy, for example SchNet and MEGNET. The graph incorporates information regarding the numbers of, types and properties of atoms; the types of bonds; and bond angles. They key to the accuracy in multi-fidelity methods, however, is the incorporation of low-fidelity output to learn the high-fidelity equivalent, in this case by learning their difference. Semi-supervised learning is employed to allow for different numbers of low and high-fidelity training points, by using an additional GCN-based low-fidelity map to predict high fidelity outputs. It is shown on 4 different data sets that a significant (at least one order of magnitude) increase in accuracy is obtained, using one to two orders of magnitude fewer low and high fidelity training points. One of the data sets is developed in this work, pertaining to 1000 simulations of quinone molecules (up to 24 atoms) at 5 different levels of fidelity, furnishing the energy, dipole moment and HOMO/LUMO.

Keywords: .materials screening, computational materials, machine learning, multi-fidelity, graph convolutional network, semi-supervised learning

Procedia PDF Downloads 33

Authors: René A. Sørensen, Søren Skovsen, Peter Christiansen, Henrik Karstoft

Abstract:

Autonomous agricultural machines act in stochastic surroundings and therefore, must be able to perceive the surroundings in real time. This perception can be achieved using image sensors combined with advanced machine learning, in particular Deep Learning. Deep convolutional neural networks excel in labeling and perceiving color images and since the cost of high-quality RGB-cameras is low, the hardware cost of good perception depends heavily on memory and computation power. This paper investigates the possibility of designing lightweight convolutional neural networks for semantic segmentation (pixel wise classification) with reduced hardware requirements, to allow for embedded usage in autonomous agricultural machines. Using compression techniques, a lightweight convolutional neural network is designed to perform real-time semantic segmentation on an embedded platform. The network is trained on two large datasets, ImageNet and Pascal Context, to recognize up to 400 individual classes. The 400 classes are remapped into agricultural superclasses (e.g. human, animal, sky, road, field, shelterbelt and obstacle) and the ability to provide accurate real-time perception of agricultural surroundings is studied. The network is applied to the case of autonomous grass mowing using the NVIDIA Tegra X1 embedded platform. Feeding case-specific images to the network results in a fully segmented map of the superclasses in the image. As the network is still being designed and optimized, only a qualitative analysis of the method is complete at the abstract submission deadline. Proceeding this deadline, the finalized design is quantitatively evaluated on 20 annotated grass mowing images. Lightweight convolutional neural networks for semantic segmentation can be implemented on an embedded platform and show competitive performance with regards to accuracy and speed. It is feasible to provide cost-efficient perceptive capabilities related to semantic segmentation for autonomous agricultural machines.

Keywords: autonomous agricultural machines, deep learning, safety, visual perception

Procedia PDF Downloads 392

Authors: Hee-Jae Suh, Kyung-Su Kim, Min-Ji Kim, Yeon-Seong Jeong, Ok-Hwan Lee, Jae-Wook Shin, Hyang-Sook Chun, Chan Lee

Abstract:

Canthaxanthin is keto-carotenoid produced from beta-carotene and it has been approved to be used in many countries as a food coloring agent. Canthaxanthin has been analyzed using High Performance Liquid Chromatography (HPLC) system with various ways of pretreatment methods. Four official methods for verification of canthaxanthin at FSA (UK), AOAC (US), EFSA (EU) and MHLW (Japan) were compared to improve its analytical and the pretreatment method. The Linearity, the limit of detection (LOD), the limit of quantification (LOQ), the accuracy, the precision and the recovery ratio were determined from each method with modification in pretreatment method. All HPLC methods exhibited correlation coefficients of calibration curves for canthaxanthin as 0.9999. The analysis methods from FSA, AOAC, and MLHW showed the LOD of 0.395 ppm, 0.105 ppm, and 0.084 ppm, and the LOQ of 1.196 ppm, 0.318 ppm, 0.254 ppm, respectively. Among tested methods, HPLC method of MHLW with modification in pretreatments was finally selected for the analysis of canthaxanthin in lab, because it exhibited the resolution factor of 4.0 and the selectivity of 1.30. This analysis method showed a correlation coefficients value of 0.9999 and the lowest LOD and LOQ. Furthermore, the precision ratio was lower than 1 and the accuracy was almost 100%. The method presented the recovery ratio of 90-110% with modification in pretreatment method. The cross-validation of coefficient variation was 5 or less among tested three institutions in Korea.

Keywords: analytic method, canthaxanthin, food colorants, pretreatment method

Procedia PDF Downloads 678