Search results for: supervised leaning
Commenced in January 2007
Frequency: Monthly
Edition: International
Paper Count: 398

Search results for: supervised leaning

248 Syndromic Surveillance Framework Using Tweets Data Analytics

Authors: David Ming Liu, Benjamin Hirsch, Bashir Aden

Abstract:

Syndromic surveillance is to detect or predict disease outbreaks through the analysis of medical sources of data. Using social media data like tweets to do syndromic surveillance becomes more and more popular with the aid of open platform to collect data and the advantage of microblogging text and mobile geographic location features. In this paper, a Syndromic Surveillance Framework is presented with machine learning kernel using tweets data analytics. Influenza and the three cities Abu Dhabi, Al Ain and Dubai of United Arabic Emirates are used as the test disease and trial areas. Hospital cases data provided by the Health Authority of Abu Dhabi (HAAD) are used for the correlation purpose. In our model, Latent Dirichlet allocation (LDA) engine is adapted to do supervised learning classification and N-Fold cross validation confusion matrix are given as the simulation results with overall system recall 85.595% performance achieved.

Keywords: Syndromic surveillance, Tweets, Machine Learning, data mining, Latent Dirichlet allocation (LDA), Influenza

Procedia PDF Downloads 91
247 Generating Music with More Refined Emotions

Authors: Shao-Di Feng, Von-Wun Soo

Abstract:

To generate symbolic music with specific emotions is a challenging task due to symbolic music datasets that have emotion labels are scarce and incomplete. This research aims to generate more refined emotions based on the training datasets that are only labeled with four quadrants in Russel’s 2D emotion model. We focus on the theory of Music Fadernet and map arousal and valence to the low-level attributes, and build a symbolic music generation model by combining transformer and GM-VAE. We adopt an in-attention mechanism for the model and improve it by allowing modulation by conditional information. And we show the music generation model could control the generation of music according to the emotions specified by users in terms of high-level linguistic expression and by manipulating their corresponding low-level musical attributes. Finally, we evaluate the model performance using a pre-trained emotion classifier against a pop piano midi dataset called EMOPIA, and by subjective listening evaluation, we demonstrate that the model could generate music with more refined emotions correctly.

Keywords: music generation, music emotion controlling, deep learning, semi-supervised learning

Procedia PDF Downloads 63
246 Exploring the Types of Infants and Toddlers' Reading Responses in Nursery Centers: A Qualitative Study

Authors: Ming Fang Hsieh

Abstract:

The purpose of this study was to investigate the reading responses of infants and toddlers across different contexts in nursery centers. The study adopted Sipe’s framework for children’s literacy education to explore the reading behavior of infants and toddlers. The study was conducted at two nurseries. The sample comprised 46 infants and toddlers and 6 caregivers. The methods of data collection included observation of various reading activities, including shared reading in a group, one-on-one reading, and unstructured reading activities, as well as interviews with caregivers. The data obtained through observations and interviews were transcribed and analyzed. The caregivers and the children’s parents signed an informed consent form before the start of the study. There was no risk anticipated during the course of the study. The analysis revealed five types of reading responses exhibited by the infants and toddlers: (1) linguistic- verbally responding to reading, repeating vocabulary, and answering questions; (2) affective- concentrating on reading or requesting for repeated reading, leaning on books, and gazing at caregivers; (3) explosive- children under 18 months were observed manipulating books through their bodies or different movements like flipping, rotating, or tapping on books; (4) social- during unstructured reading context, children were seen interacting with peers or following the rules of reading, sitting properly, and choosing one book at a time; and (5) distracted responses- paying attention to something else instead of reading, walking around, and playing, which was usually observed during shared reading in a group. The study concluded that children’s distraction and explosive reading behaviors may be a part of the process of their emergent reading behavior. As children develop, they demonstrate an increase in verbal responses, improved concentration, and better behavior. The study suggests that adults should continue to provide appropriate reading opportunities beginning from infancy to nurture children’s reading behaviors.

Keywords: reading response, infants and toddlers, early reading, picture books

Procedia PDF Downloads 83
245 Improving the Performance of Back-Propagation Training Algorithm by Using ANN

Authors: Vishnu Pratap Singh Kirar

Abstract:

Artificial Neural Network (ANN) can be trained using backpropagation (BP). It is the most widely used algorithm for supervised learning with multi-layered feed-forward networks. Efficient learning by the BP algorithm is required for many practical applications. The BP algorithm calculates the weight changes of artificial neural networks, and a common approach is to use a two-term algorithm consisting of a learning rate (LR) and a momentum factor (MF). The major drawbacks of the two-term BP learning algorithm are the problems of local minima and slow convergence speeds, which limit the scope for real-time applications. Recently the addition of an extra term, called a proportional factor (PF), to the two-term BP algorithm was proposed. The third increases the speed of the BP algorithm. However, the PF term also reduces the convergence of the BP algorithm, and criteria for evaluating convergence are required to facilitate the application of the three terms BP algorithm. Although these two seem to be closely related, as described later, we summarize various improvements to overcome the drawbacks. Here we compare the different methods of convergence of the new three-term BP algorithm.

Keywords: neural network, backpropagation, local minima, fast convergence rate

Procedia PDF Downloads 473
244 Passenger Movement Pattern during Ship Evacuation Considering the Combined Effect of Ship Heeling and Trim

Authors: Jinlu Sun, Shouxiang Lu, Siuming Lo

Abstract:

Large passenger ship, especially luxury cruise, is one of the most prevalent means of marine transportation and tourism nowadays. In case of an accident, an effective evacuation would be the ultimate way to minimize the consequence. Ship heeling and trim has a considerable influence on passenger walking speed and posture during ship evacuation. To investigate passenger movement pattern under the combined effect of ship heeling and trim, a ship corridor simulator was developed. Both fast and freely individual walking experiments by male and female experimental subjects under heeling and trim conditions were conducted and recorded therein. It is found that routes of experimental subjects would change due to the heeling and trim angles, although they always walk along the right side because of cultural factors. Experimental subjects would also change their posture to adapt the combined heeling and trim conditions, such as leaning forward, adopting larger arm swaying, shorter and more frequent steps. While for individual walking speed, the speed would decrease with the increasing heeling and trim angles. But the maximum individual walking speed is achieved at heeling angle of 0° with trim angle ranging from -15° to -5 °, instead of on level ground, which may be attributable to the effect of the gravitational acceleration. Female is approximately 10% slower than male due to the discrepancy in physical quality. Besides, individual walking speed shows similar trends in both fast and freely walking modes, and the speed value in freely walking mode is about 78% of that in fast walking mode under each experimental condition. Furthermore, to designate the movement pattern of passengers in heeling and trim conditions, a model of the walking speed reduction was proposed. This work would provide guidance on the development of evacuation models and the design of evacuation facilities on board.

Keywords: evacuation, heeling, individual walking speed, ship corridor simulator, trim

Procedia PDF Downloads 234
243 An Assessment of the Usage of Learner Centred Methods among Student Teachers of Federal College of Education Kontagora

Authors: Sadiq Habiba Alhaji

Abstract:

This is a descriptive survey design intended to determine the level of usage of the learner centred methods by student teachers of Federal College of Education Kontagora, Niger State, Nigeria. The study was guided by two null hypotheses formulated by the researcher. The population of the study are students of Federal College of Education, Kontagora. The Target Population consisted of one hundred Teaching practice students drawn from sciences, Arts, and humanities who were posted to various schools practicing different teaching methods. The student teachers were supervised using the checklist designed by the researcher to determine their level of usage of learner centred methods. Data collected was analysed using t test of independent variables. It was recommended that pre service and in service teachers should be equipped with the skills of using learner centred methods.

Keywords: assessment, usage, learner centred, methods, student teachers

Procedia PDF Downloads 67
242 Classification of IoT Traffic Security Attacks Using Deep Learning

Authors: Anum Ali, Kashaf ad Dooja, Asif Saleem

Abstract:

The future smart cities trend will be towards Internet of Things (IoT); IoT creates dynamic connections in a ubiquitous manner. Smart cities offer ease and flexibility for daily life matters. By using small devices that are connected to cloud servers based on IoT, network traffic between these devices is growing exponentially, whose security is a concerned issue, since ratio of cyber attack may make the network traffic vulnerable. This paper discusses the latest machine learning approaches in related work further to tackle the increasing rate of cyber attacks, machine learning algorithm is applied to IoT-based network traffic data. The proposed algorithm train itself on data and identify different sections of devices interaction by using supervised learning which is considered as a classifier related to a specific IoT device class. The simulation results clearly identify the attacks and produce fewer false detections.

Keywords: IoT, traffic security, deep learning, classification

Procedia PDF Downloads 130
241 Using Classifiers to Predict Student Outcome at Higher Institute of Telecommunication

Authors: Fuad M. Alkoot

Abstract:

We aim at highlighting the benefits of classifier systems especially in supporting educational management decisions. The paper aims at using classifiers in an educational application where an outcome is predicted based on given input parameters that represent various conditions at the institute. We present a classifier system that is designed using a limited training set with data for only one semester. The achieved system is able to reach at previously known outcomes accurately. It is also tested on new input parameters representing variations of input conditions to see its prediction on the possible outcome value. Given the supervised expectation of the outcome for the new input we find the system is able to predict the correct outcome. Experiments were conducted on one semester data from two departments only, Switching and Mathematics. Future work on other departments with larger training sets and wider input variations will show additional benefits of classifier systems in supporting the management decisions at an educational institute.

Keywords: machine learning, pattern recognition, classifier design, educational management, outcome estimation

Procedia PDF Downloads 258
240 Adaption Model for Building Agile Pronunciation Dictionaries Using Phonemic Distance Measurements

Authors: Akella Amarendra Babu, Rama Devi Yellasiri, Natukula Sainath

Abstract:

Where human beings can easily learn and adopt pronunciation variations, machines need training before put into use. Also humans keep minimum vocabulary and their pronunciation variations are stored in front-end of their memory for ready reference, while machines keep the entire pronunciation dictionary for ready reference. Supervised methods are used for preparation of pronunciation dictionaries which take large amounts of manual effort, cost, time and are not suitable for real time use. This paper presents an unsupervised adaptation model for building agile and dynamic pronunciation dictionaries online. These methods mimic human approach in learning the new pronunciations in real time. A new algorithm for measuring sound distances called Dynamic Phone Warping is presented and tested. Performance of the system is measured using an adaptation model and the precision metrics is found to be better than 86 percent.

Keywords: pronunciation variations, dynamic programming, machine learning, natural language processing

Procedia PDF Downloads 150
239 Automated Detection of Women Dehumanization in English Text

Authors: Maha Wiss, Wael Khreich

Abstract:

Animals, objects, foods, plants, and other non-human terms are commonly used as a source of metaphors to describe females in formal and slang language. Comparing women to non-human items not only reflects cultural views that might conceptualize women as subordinates or in a lower position than humans, yet it conveys this degradation to the listeners. Moreover, the dehumanizing representation of females in the language normalizes the derogation and even encourages sexism and aggressiveness against women. Although dehumanization has been a popular research topic for decades, according to our knowledge, no studies have linked women's dehumanizing language to the machine learning field. Therefore, we introduce our research work as one of the first attempts to create a tool for the automated detection of the dehumanizing depiction of females in English texts. We also present the first labeled dataset on the charted topic, which is used for training supervised machine learning algorithms to build an accurate classification model. The importance of this work is that it accomplishes the first step toward mitigating dehumanizing language against females.

Keywords: gender bias, machine learning, NLP, women dehumanization

Procedia PDF Downloads 60
238 On-Line Data-Driven Multivariate Statistical Prediction Approach to Production Monitoring

Authors: Hyun-Woo Cho

Abstract:

Detection of incipient abnormal events in production processes is important to improve safety and reliability of manufacturing operations and reduce losses caused by failures. The construction of calibration models for predicting faulty conditions is quite essential in making decisions on when to perform preventive maintenance. This paper presents a multivariate calibration monitoring approach based on the statistical analysis of process measurement data. The calibration model is used to predict faulty conditions from historical reference data. This approach utilizes variable selection techniques, and the predictive performance of several prediction methods are evaluated using real data. The results shows that the calibration model based on supervised probabilistic model yielded best performance in this work. By adopting a proper variable selection scheme in calibration models, the prediction performance can be improved by excluding non-informative variables from their model building steps.

Keywords: calibration model, monitoring, quality improvement, feature selection

Procedia PDF Downloads 339
237 Monitoring the Rate of Expansion of Agricultural Fields in Mwekera Forest Reserve Using Remote Sensing and Geographic Information Systems

Authors: K. Kanja, M. Mweemba, K. Malungwa

Abstract:

Due to the rampant population growth coupled with retrenchments currently going on in the Copper mines in Zambia, a number of people are resorting to land clearing for agriculture, illegal settlements as well as charcoal production among other vices. This study aims at assessing the rate of expansion of agricultural fields and illegal settlements in protected areas using remote sensing and Geographic Information System. Zambia’s Mwekera National Forest Reserve was used as a case study. Iterative Self-Organizing Data Analysis Technique (ISODATA), as well as maximum likelihood, supervised classification on four Landsat images as well as an accuracy assessment of the classifications was performed. Over the period under observation, results indicate annual percentage changes to be -0.03, -0.49 and 1.26 for agriculture, forests and settlement respectively indicating a higher conversion of forests into human settlements and agriculture.

Keywords: geographic information system, land cover change, Landsat TM and ETM+, Mwekera forest reserve, remote sensing

Procedia PDF Downloads 125
236 Land Use Change Detection Using Remote Sensing and GIS

Authors: Naser Ahmadi Sani, Karim Solaimani, Lida Razaghnia, Jalal Zandi

Abstract:

In recent decades, rapid and incorrect changes in land-use have been associated with consequences such as natural resources degradation and environmental pollution. Detecting changes in land-use is one of the tools for natural resource management and assessment of changes in ecosystems. The target of this research is studying the land-use changes in Haraz basin with an area of 677000 hectares in a 15 years period (1996 to 2011) using LANDSAT data. Therefore, the quality of the images was first evaluated. Various enhancement methods for creating synthetic bonds were used in the analysis. Separate training sites were selected for each image. Then the images of each period were classified in 9 classes using supervised classification method and the maximum likelihood algorithm. Finally, the changes were extracted in GIS environment. The results showed that these changes are an alarm for the HARAZ basin status in future. The reason is that 27% of the area has been changed, which is related to changing the range lands to bare land and dry farming and also changing the dense forest to sparse forest, horticulture, farming land and residential area.

Keywords: Haraz basin, change detection, land-use, satellite data

Procedia PDF Downloads 394
235 Utilizing Bario Rice, a Natural Red-Pigmented Rice from Sarawak, Malaysia, in the Development of Gluten-Free Bread

Authors: Macdalyna Esther Ronie, Hasmadi Mamat, Ahmad Hazim Abdul Aziz, Mohamad Khairi Zainol

Abstract:

Current trends in gluten-free food development are increasingly leaning towards the utilization of pigmented rice flour, with a particular focus on Bario Merah Sederhana (BMS), a red-pigmented rice native to Sarawak, Malaysia. This study delves into the evaluation of the nutritional, textural, and sensory attributes of gluten-free rice bread produced from a blend of BMS rice flour and potato starch. The resulting samples are denoted as F1 (100% BMS rice flour), F2 (90% BMS rice flour and 10% potato starch), F3 (80% BMS rice flour and 20% potato starch), and F4 (70% BMS rice flour and 30% potato starch). Comparatively, these gluten-free rice bread formulations exhibit higher levels of ash and crude fiber, along with lower carbohydrate content when juxtaposed with conventional wheat bread. Notably, the crude protein content of the rice bread diminishes significantly (p<0.05) as the proportion of rice flour decreases, primarily due to the higher protein content found in wheat flour. The crumb of the rice bread appears darker owing to the red pigment in the rice flour, while the crust is lighter than that of the control sample, possibly attributable to a reduced Maillard reaction. Among the various rice bread formulations, F4 stands out with the least dough and bread hardness, accompanied by the highest levels of stickiness and springiness in both dough and bread, respectively. In sensory evaluations, wheat bread garners the highest rating (p<0.05). However, within the realm of rice breads, F4 emerges as a viable and acceptable formulation, as indicated by its commendable scores in color (7.03), flavor (5.73), texture (6.03), and overall acceptability (6.18). These findings underscore the potential of BMS in the creation of gluten-free rice breads, with the formulation consisting of 70% rice flour and 30% potato starch emerging as a well-received and suitable option.

Keywords: gluten-free bread, bario rice, proximate composition, sensory evaluation

Procedia PDF Downloads 213
234 Modified Naive Bayes-Based Prediction Modeling for Crop Yield Prediction

Authors: Kefaya Qaddoum

Abstract:

Most of greenhouse growers desire a determined amount of yields in order to accurately meet market requirements. The purpose of this paper is to model a simple but often satisfactory supervised classification method. The original naive Bayes have a serious weakness, which is producing redundant predictors. In this paper, utilized regularization technique was used to obtain a computationally efficient classifier based on naive Bayes. The suggested construction, utilized L1-penalty, is capable of clearing redundant predictors, where a modification of the LARS algorithm is devised to solve this problem, making this method applicable to a wide range of data. In the experimental section, a study conducted to examine the effect of redundant and irrelevant predictors, and test the method on WSG data set for tomato yields, where there are many more predictors than data, and the urge need to predict weekly yield is the goal of this approach. Finally, the modified approach is compared with several naive Bayes variants and other classification algorithms (SVM and kNN), and is shown to be fairly good.

Keywords: tomato yield prediction, naive Bayes, redundancy, WSG

Procedia PDF Downloads 213
233 Loneliness and Depression in Relation to Latchkey Situation

Authors: Samaneh Sadat Fattahi Massoom, Hossein Salimi Bajestani

Abstract:

The study examines loneliness and depression in students who regularly care for themselves after school (latchkey students) in Mashhad and compares them with parent supervised students using a causal-comparative research method. The 270 participants, aged 7 -13, were selected using convenience and cluster random-assignment sampling. Independent t-test results showed significant differences between loneliness (-4.32, p ≤ 0.05) and depression (-3.02, p ≤0.05) among latchkey and non-latchkey students. Using the Pearson correlation test, significant correlation between depression and loneliness among latchkey students was also discovered (r=0.59, p ≤ 0.05). However, regarding non latchkey students, no significant difference between loneliness and depression was observed (r= 0.02. p ≥ 0.05). Multiple regression results also showed that depression variance can be determined by gender (22%) and loneliness (34%). The findings of this study, specifically the significant difference between latchkey and non-latchkey children regarding feelings of loneliness and depression, carries clear implications for parents. It can be concluded that mothers who spend most of their time working out of the house and devoid their children of their presence in the home may cause some form of mental distress like loneliness and depression. Moreover, gender differences affect the degree of these psychological disorders.

Keywords: loneliness, depression, self-care students, latchkey and non-latchkey students, gender

Procedia PDF Downloads 385
232 A Reflective Investigation on the Course Design and Coaching Strategy for Creating a Trans-Disciplinary Leaning Environment

Authors: Min-Feng Hsieh

Abstract:

Nowadays, we are facing a highly competitive environment in which the situation for survival has come even more critical than ever before. The challenge we will be confronted with is no longer can be dealt with the single system of knowledge. The abilities we urgently need to acquire is something that can lead us to cross over the boundaries between different disciplines and take us to a neutral ground that gathers and integrates powers and intelligence that surrounds us. This paper aims at discussing how a trans-disciplinary design course organized by the College of Design at Chaoyang University can react to this modern challenge. By orchestrating an experimental course format and by developing a series of coaching strategies, a trans-disciplinary learning environment has been created and practiced in which students selected from five different departments, including Architecture, Interior Design, Visual Design, Industrial Design, Landscape and Urban Design, are encouraged to think outside their familiar knowledge pool and to learn with/from each other. In the course of implementing this program, a parallel research has been conducted alongside by adopting the theory and principles of Action Research which is a research methodology that can provide the course organizer emergent, responsive, action-oriented, participative and critically reflective insights for the immediate changes and amendments in order to improve the effect of teaching and learning experience. In the conclusion, how the learning and teaching experience of this trans-disciplinary design studio can offer us some observation that can help us reflect upon the constraints and division caused by the subject base curriculum will be pointed out. A series of concepts for course design and teaching strategies developed and implemented in this trans-disciplinary course are to be introduced as a way to promote learners’ self-motivated, collaborative, cross-disciplinary and student-centered learning skills. The outcome of this experimental course can exemplify an alternative approach that we could adopt in pursuing a remedy for dealing with the problematic issues of the current educational practice.

Keywords: course design, coaching strategy, subject base curriculum, trans-disciplinary

Procedia PDF Downloads 185
231 Multilabel Classification with Neural Network Ensemble Method

Authors: Sezin Ekşioğlu

Abstract:

Multilabel classification has a huge importance for several applications, it is also a challenging research topic. It is a kind of supervised learning that contains binary targets. The distance between multilabel and binary classification is having more than one class in multilabel classification problems. Features can belong to one class or many classes. There exists a wide range of applications for multi label prediction such as image labeling, text categorization, gene functionality. Even though features are classified in many classes, they may not always be properly classified. There are many ensemble methods for the classification. However, most of the researchers have been concerned about better multilabel methods. Especially little ones focus on both efficiency of classifiers and pairwise relationships at the same time in order to implement better multilabel classification. In this paper, we worked on modified ensemble methods by getting benefit from k-Nearest Neighbors and neural network structure to address issues within a beneficial way and to get better impacts from the multilabel classification. Publicly available datasets (yeast, emotion, scene and birds) are performed to demonstrate the developed algorithm efficiency and the technique is measured by accuracy, F1 score and hamming loss metrics. Our algorithm boosts benchmarks for each datasets with different metrics.

Keywords: multilabel, classification, neural network, KNN

Procedia PDF Downloads 132
230 Comparing the Effect of Group Education and Multimedia Software on Knowledge, Attitude and Self-Efficacy Mothers about of Sexual Health Education to the Boys of between 12-14 Years Old

Authors: Mirzaii Khadigeh

Abstract:

Background and objectives: Sexual health education is an important part of health promotion services. The major role of sex education is on mothers’ shoulders. So, they have to be equipped with enough knowledge, attitude and self-efficacy for teens’ education. The present study aimed to determine the effect of team-learning and multimedia software on mothers’ knowledge, attitudes and self-efficacy in sexual health education to 12-14-year-old sons in Mashhad in 1395. Materials and methods: In this research, two experimental and one control group were employed using random sampling, which was done on 132 mothers of high school pupils. They were randomly assigned into experimental and control groups. The data were collected using demographic information and a researcher-constructed questionnaire to investigate the mothers’ knowledge, attitude, and self-efficacy and DASS21(The Depression, Anxiety and Stress Scale). They were run after confirming their reliability and validity. Intervention for the multimedia group was in the form of four CDs- each for 45 minutes- that were given to the mothers each week. At the end of the fourth week, a question-and-answer session was administered for 60 minutes. The team-learning group received intervention once a week (totally four weeks). Two weeks later, the data were collected and analyzed via Chi-square, Fisher, Kruskal-Wallis and ANOVA. Findings: Knowledge, attitude and self-efficacy of mothers in sexual health before the intervention did not have any significant differences (p >0.05). At the end of the study, the difference between the scores of the knowledge, attitude and self-efficacy in the three groups was meaningfully different (p < 0/001), but the difference between the two groups of multimedia and team-learning was not significant (p> 0.05 ). Discussion and conclusion: The result reported the efficacy of both team-leaning and multimedia software, which implies that the multimedia software training method was as effective as team-learning training one on the knowledge, attitude and self-efficacy of mothers. But, the multimedia training method is highly advised due to its availability, flexibility, and interest.

Keywords: training one on the knowledge, attitude, self-efficacy of mothers, boys

Procedia PDF Downloads 152
229 Efficiency of Information Technology Based Learning and Teaching in Higher Educations

Authors: Mahalingam Palaniandi

Abstract:

Higher education plays vital role in the nation building process for a country and the rest of world. The higher education sector develops the change-agents for the various fields which will help the human-kind wheel to run further. Conventional and traditional class-room based learning and teaching was followed in many decades which is one-to-one and one-to-many. In a way, these are simplest form of learners to be assembled in a class room wherein the teacher used the blackboard to demonstrate the theory and laboratories used for practical. As the technology evolved tremendously for the last 40 years, the teaching and learning environment changed slowly, wherein, the learning community will be anywhere in the world and teacher deliver the content through internet based tools such as video conferencing, web based conferencing tools or E-learning platforms such as Blackboard or noodle. Present day, the mobile technologies plays an important tool to deliver the teaching content on-the-go. Both PC based and mobile based learning technology brought the learning and teaching community together in various aspects. However, as the learning technology also brought various hurdles for learning processes such as plagiarism and not using the reference books entirely as most of the students wants the information instantaneously using internet without actually going to the library to take the notes from the millions of the books which are not available online as e-books which result lack of fundamental knowledge of the concepts complex theories. However, technology is inseparable in human life, now-a-days and every part of it contains piece of information technology right from computers to home appliances. To make use of the IT based learning and teaching at most efficiency, we should have a proper framework and recommendations laid to the learning community in order to derive the maximum efficiency from the IT based teaching and leaning. This paper discusses various IT based tools available for the learning community, efficiency from its usage and recommendations for the suitable framework that needs to be implemented at higher education institutions which makes the learners stronger in both theory as well as real-time knowledge of their studies that is going to be used in their future for the better world.

Keywords: higher education, e-learning, teaching learning, eLearning tools

Procedia PDF Downloads 403
228 Deep learning with Noisy Labels : Learning True Labels as Discrete Latent Variable

Authors: Azeddine El-Hassouny, Chandrashekhar Meshram, Geraldin Nanfack

Abstract:

In recent years, learning from data with noisy labels (Label Noise) has been a major concern in supervised learning. This problem has become even more worrying in Deep Learning, where the generalization capabilities have been questioned lately. Indeed, deep learning requires a large amount of data that is generally collected by search engines, which frequently return data with unreliable labels. In this paper, we investigate the Label Noise in Deep Learning using variational inference. Our contributions are : (1) exploiting Label Noise concept where the true labels are learnt using reparameterization variational inference, while observed labels are learnt discriminatively. (2) the noise transition matrix is learnt during the training without any particular process, neither heuristic nor preliminary phases. The theoretical results shows how true label distribution can be learned by variational inference in any discriminate neural network, and the effectiveness of our approach is proved in several target datasets, such as MNIST and CIFAR32.

Keywords: label noise, deep learning, discrete latent variable, variational inference, MNIST, CIFAR32

Procedia PDF Downloads 100
227 Supervised-Component-Based Generalised Linear Regression with Multiple Explanatory Blocks: THEME-SCGLR

Authors: Bry X., Trottier C., Mortier F., Cornu G., Verron T.

Abstract:

We address component-based regularization of a Multivariate Generalized Linear Model (MGLM). A set of random responses Y is assumed to depend, through a GLM, on a set X of explanatory variables, as well as on a set T of additional covariates. X is partitioned into R conceptually homogeneous blocks X1, ... , XR , viewed as explanatory themes. Variables in each Xr are assumed many and redundant. Thus, Generalised Linear Regression (GLR) demands regularization with respect to each Xr. By contrast, variables in T are assumed selected so as to demand no regularization. Regularization is performed searching each Xr for an appropriate number of orthogonal components that both contribute to model Y and capture relevant structural information in Xr. We propose a very general criterion to measure structural relevance (SR) of a component in a block, and show how to take SR into account within a Fisher-scoring-type algorithm in order to estimate the model. We show how to deal with mixed-type explanatory variables. The method, named THEME-SCGLR, is tested on simulated data.

Keywords: Component-Model, Fisher Scoring Algorithm, GLM, PLS Regression, SCGLR, SEER, THEME

Procedia PDF Downloads 379
226 An Integrated Approach to Find the Effect of Strain Rate on Ultimate Tensile Strength of Randomly Oriented Short Glass Fiber Composite in Combination with Artificial Neural Network

Authors: Sharad Shrivastava, Arun Jalan

Abstract:

In this study tensile testing was performed on randomly oriented short glass fiber/epoxy resin composite specimens which were prepared using hand lay-up method. Samples were tested over a wide range of strain rate/loading rate from 2mm/min to 40mm/min to see the effect on ultimate tensile strength of the composite. A multi layered 'back propagation artificial neural network of supervised learning type' was used to analyze and predict the tensile properties with strain rate and temperature as given input and output as UTS to predict. Various network structures were designed and investigated with varying parameters and network sizes, and an optimized network structure was proposed to predict the UTS of short glass fiber/epoxy resin composite specimens with reasonably good accuracy.

Keywords: glass fiber composite, mechanical properties, strain rate, artificial neural network

Procedia PDF Downloads 419
225 Automatic Classification of Periodic Heart Sounds Using Convolutional Neural Network

Authors: Jia Xin Low, Keng Wah Choo

Abstract:

This paper presents an automatic normal and abnormal heart sound classification model developed based on deep learning algorithm. MITHSDB heart sounds datasets obtained from the 2016 PhysioNet/Computing in Cardiology Challenge database were used in this research with the assumption that the electrocardiograms (ECG) were recorded simultaneously with the heart sounds (phonocardiogram, PCG). The PCG time series are segmented per heart beat, and each sub-segment is converted to form a square intensity matrix, and classified using convolutional neural network (CNN) models. This approach removes the need to provide classification features for the supervised machine learning algorithm. Instead, the features are determined automatically through training, from the time series provided. The result proves that the prediction model is able to provide reasonable and comparable classification accuracy despite simple implementation. This approach can be used for real-time classification of heart sounds in Internet of Medical Things (IoMT), e.g. remote monitoring applications of PCG signal.

Keywords: convolutional neural network, discrete wavelet transform, deep learning, heart sound classification

Procedia PDF Downloads 326
224 Methodology of Automation and Supervisory Control and Data Acquisition for Restructuring Industrial Systems

Authors: Lakhoua Najeh

Abstract:

Introduction: In most situations, an industrial system already existing, conditioned by its history, its culture and its context are in difficulty facing the necessity to restructure itself in an organizational and technological environment in perpetual evolution. This is why all operations of restructuring first of all require a diagnosis based on a functional analysis. After a presentation of the functionality of a supervisory system for complex processes, we present the concepts of industrial automation and supervisory control and data acquisition (SCADA). Methods: This global analysis exploits the various available documents on the one hand and takes on the other hand in consideration the various testimonies through investigations, the interviews or the collective workshops; otherwise, it also takes observations through visits as a basis and even of the specific operations. The exploitation of this diagnosis enables us to elaborate the project of restructuring thereafter. Leaving from the system analysis for the restructuring of industrial systems, and after a technical diagnosis based on visits, an analysis of the various technical documents and management as well as on targeted interviews, a focusing retailing the various levels of analysis has been done according a general methodology. Results: The methodology adopted in order to contribute to the restructuring of industrial systems by its participative and systemic character and leaning on a large consultation a lot of human resources that of the documentary resources, various innovating actions has been proposed. These actions appear in the setting of the TQM gait requiring applicable parameter quantification and a treatment valorising some information. The new management environment will enable us to institute an information and communication system possibility of migration toward an ERP system. Conclusion: Technological advancements in process monitoring, control and industrial automation over the past decades have contributed greatly to improve the productivity of virtually all industrial systems throughout the world. This paper tries to identify the principles characteristics of a process monitoring, control and industrial automation in order to provide tools to help in the decision-making process.

Keywords: automation, supervision, SCADA, TQM

Procedia PDF Downloads 147
223 Selection of Appropriate Classification Technique for Lithological Mapping of Gali Jagir Area, Pakistan

Authors: Khunsa Fatima, Umar K. Khattak, Allah Bakhsh Kausar

Abstract:

Satellite images interpretation and analysis assist geologists by providing valuable information about geology and minerals of an area to be surveyed. A test site in Fatejang of district Attock has been studied using Landsat ETM+ and ASTER satellite images for lithological mapping. Five different supervised image classification techniques namely maximum likelihood, parallelepiped, minimum distance to mean, mahalanobis distance and spectral angle mapper have been performed on both satellite data images to find out the suitable classification technique for lithological mapping in the study area. Results of these five image classification techniques were compared with the geological map produced by Geological Survey of Pakistan. The result of maximum likelihood classification technique applied on ASTER satellite image has the highest correlation of 0.66 with the geological map. Field observations and XRD spectra of field samples also verified the results. A lithological map was then prepared based on the maximum likelihood classification of ASTER satellite image.

Keywords: ASTER, Landsat-ETM+, satellite, image classification

Procedia PDF Downloads 368
222 A Supervised Approach for Word Sense Disambiguation Based on Arabic Diacritics

Authors: Alaa Alrakaf, Sk. Md. Mizanur Rahman

Abstract:

Since the last two decades’ Arabic natural language processing (ANLP) has become increasingly much more important. One of the key issues related to ANLP is ambiguity. In Arabic language different pronunciation of one word may have a different meaning. Furthermore, ambiguity also has an impact on the effectiveness and efficiency of Machine Translation (MT). The issue of ambiguity has limited the usefulness and accuracy of the translation from Arabic to English. The lack of Arabic resources makes ambiguity problem more complicated. Additionally, the orthographic level of representation cannot specify the exact meaning of the word. This paper looked at the diacritics of Arabic language and used them to disambiguate a word. The proposed approach of word sense disambiguation used Diacritizer application to Diacritize Arabic text then found the most accurate sense of an ambiguous word using Naïve Bayes Classifier. Our Experimental study proves that using Arabic Diacritics with Naïve Bayes Classifier enhances the accuracy of choosing the appropriate sense by 23% and also decreases the ambiguity in machine translation.

Keywords: Arabic natural language processing, machine learning, machine translation, Naive bayes classifier, word sense disambiguation

Procedia PDF Downloads 335
221 Microfungi on Sandy Beaches: Potential Threats for People Enjoying Lakeside Recreation

Authors: Tomasz Balabanski, Anna Biedunkiewicz

Abstract:

Research on basic bacteriological and physicochemical parameters conducted by state institutions (Provincial Sanitary and Epidemiological Station and District Sanitary and Epidemiological Station) are limited to bathing waters under constant sanitary and epidemiological supervision. Unfortunately, no routine or monitoring tests are carried out for the presence of microfungi. This also applies to beach sand used for recreational purposes. The purpose of the planned own research was to determine the diversity of the mycobiota present on supervised and unsupervised sandy beaches, on the shores of lakes, of municipal baths used for recreation. The research material consisted of microfungi isolated from April to October 2019 from sandy beaches of supervised and unsupervised lakes located within the administrative boundaries of the city of Olsztyn (North-Eastern Poland, Europe). Four lakes, out of the fifteen available (Tyrsko, Kortowskie, Skanda, and Ukiel), whose bathing waters are subjected to routine bacteriological tests, were selected for testing. To compare the diversity of the mycobiota composition on the surface and below the sand mixing layer, samples were taken from two depths (10 cm and 50 cm), using a soil auger. Micro-fungi from sand samples were obtained by surface inoculation on an RBC medium from the 1st dilution (1:10). After incubation at 25°C for 96-144 h, the average number of CFU/dm³ was counted. Morphologically differing yeast colonies were passaged into Sabouraud agar slants with gentamicin and incubated again. For detailed laboratory analyses, culture methods (macro- and micro-cultures) and identification methods recommended in diagnostic mycological laboratories were used. The conducted research allowed obtaining 140 yeast isolates. The total average population ranged from 1.37 × 10⁻² CFU/dm³ before the bathing season (April 2019), 1.64 × 10⁻³ CFU/dm³ in the season (May-September 2019), and 1.60 × 10⁻² CFU/dm³ after the end of the season (October 2019). More microfungi were obtained from the surface layer of sand (100 isolates) than from the deeper layer (40 isolates). Reported microfungi may circulate seasonally between individual elements of the lake ecosystem. From the sand/soil from the catchment area beaches, they can get into bathing waters, stopping periodically on the coastal phyllosphere. The sand of the beaches and the phyllosphere are a kind of filter for the water reservoir. The presence of microfungi with various pathogenicity potential in these places is of major epidemiological importance. Therefore, full monitoring of not only recreational waters but also sandy beaches should be treated as an element of constant control by appropriate supervisory institutions, allowing recreational areas for public use so that the use of these places does not involve the risk of infection. Acknowledgment: 'Development Program of the University of Warmia and Mazury in Olsztyn', POWR.03.05.00-00-Z310/17, co-financed by the European Union under the European Social Fund from the Operational Program Knowledge Education Development. Tomasz Bałabański is a recipient of a scholarship from the Programme Interdisciplinary Doctoral Studies in Biology and Biotechnology (POWR.03.05.00-00-Z310/17), which is funded by the 'European Social Fund'.

Keywords: beach, microfungi, sand, yeasts

Procedia PDF Downloads 78
220 Research on Knowledge Graph Inference Technology Based on Proximal Policy Optimization

Authors: Yihao Kuang, Bowen Ding

Abstract:

With the increasing scale and complexity of knowledge graph, modern knowledge graph contains more and more types of entity, relationship, and attribute information. Therefore, in recent years, it has been a trend for knowledge graph inference to use reinforcement learning to deal with large-scale, incomplete, and noisy knowledge graph and improve the inference effect and interpretability. The Proximal Policy Optimization (PPO) algorithm utilizes a near-end strategy optimization approach. This allows for more extensive updates of policy parameters while constraining the update extent to maintain training stability. This characteristic enables PPOs to converge to improve strategies more rapidly, often demonstrating enhanced performance early in the training process. Furthermore, PPO has the advantage of offline learning, effectively utilizing historical experience data for training and enhancing sample utilization. This means that even with limited resources, PPOs can efficiently train for reinforcement learning tasks. Based on these characteristics, this paper aims to obtain better and more efficient inference effect by introducing PPO into knowledge inference technology.

Keywords: reinforcement learning, PPO, knowledge inference, supervised learning

Procedia PDF Downloads 42
219 ANN Based Simulation of PWM Scheme for Seven Phase Voltage Source Inverter Using MATLAB/Simulink

Authors: Mohammad Arif Khan

Abstract:

This paper analyzes and presents the development of Artificial Neural Network based controller of space vector modulation (ANN-SVPWM) for a seven-phase voltage source inverter. At first, the conventional method of producing sinusoidal output voltage by utilizing six active and one zero space vectors are used to synthesize the input reference, is elaborated and then new PWM scheme called Artificial Neural Network Based PWM is presented. The ANN based controller has the advantage of the very fast implementation and analyzing the algorithms and avoids the direct computation of trigonometric and non-linear functions. The ANN controller uses the individual training strategy with the fixed weight and supervised models. A computer simulation program has been developed using Matlab/Simulink together with the neural network toolbox for training the ANN-controller. A comparison of the proposed scheme with the conventional scheme is presented based on various performance indices. Extensive Simulation results are provided to validate the findings.

Keywords: space vector PWM, total harmonic distortion, seven-phase, voltage source inverter, multi-phase, artificial neural network

Procedia PDF Downloads 435