Search results for: category learning

Authors: Mohammad Reza Karami Nejad

Abstract:

A hybrid learning automata-genetic algorithm (HLGA) is proposed to solve QoS routing optimization problem of next generation networks. The algorithm complements the advantages of the learning Automato Algorithm(LA) and Genetic Algorithm(GA). It firstly uses the good global search capability of LA to generate initial population needed by GA, then it uses GA to improve the Quality of Service(QoS) and acquiring the optimization tree through new algorithms for crossover and mutation operators which are an NP-Complete problem. In the proposed algorithm, the connectivity matrix of edges is used for genotype representation. Some novel heuristics are also proposed for mutation, crossover, and creation of random individuals. We evaluate the performance and efficiency of the proposed HLGA-based algorithm in comparison with other existing heuristic and GA-based algorithms by the result of simulation. Simulation results demonstrate that this paper proposed algorithm not only has the fast calculating speed and high accuracy but also can improve the efficiency in Next Generation Networks QoS routing. The proposed algorithm has overcome all of the previous algorithms in the literature.

Keywords: Routing, Quality of Service, Multicaset, Learning Automata, Genetic, Next Generation Networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1731

Authors: Sun-Young Jang, Sung-Ah Kim, Dongyoun Shin

Abstract:

Current real-estate value estimation, difficult for laymen, usually is performed by specialists. This paper presents an automated estimation process based on big data and machine-learning technology that calculates influences of building conditions on real-estate price measurement. The present study analyzed actual building sales sample data for Nonhyeon-dong, Gangnam-gu, Seoul, Korea, measuring the major influencing factors among the various building conditions. Further to that analysis, a prediction model was established and applied using RapidMiner Studio, a graphical user interface (GUI)-based tool for derivation of machine-learning prototypes. The prediction model is formulated by reference to previous examples. When new examples are applied, it analyses and predicts accordingly. The analysis process discerns the crucial factors effecting price increases by calculation of weighted values. The model was verified, and its accuracy determined, by comparing its predicted values with actual price increases.

Keywords: Big data, building-value analysis, machine learning, price prediction.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1156

Authors: Brandy Yee, Dianne Yee

Abstract:

Through the exploration of the lived experiences, beliefs and values of instructional leaders, teachers and students in Finland, Germany and Canada, we investigated the factors which contribute to developmentally responsive, intellectually engaging middle-level learning environments for early adolescents. Student-centred leadership dimensions, effective instructional practices and student agency were examined through the lens of current policy and research on middle-level learning environments emerging from the Canadian province of Manitoba. Consideration of these three research perspectives in the context of early adolescent learning, placed against an international backdrop, provided a previously undocumented perspective on leading, teaching and learning in the middle years. Aligning with a social constructivist, qualitative research paradigm, the study incorporated collective case study methodology, along with constructivist grounded theory methods of data analysis. Data were collected through semi-structured individual and focus group interviews and document review, as well as direct and participant observation. Three case study narratives were developed to share the rich stories of study participants, who had been selected using maximum variation and intensity sampling techniques. Interview transcript data were coded using processes from constructivist grounded theory. A cross-case analysis yielded a conceptual framework highlighting key factors that were found to be significant in the establishment of developmentally responsive, intellectually engaging middle-level learning environments. Seven core categories emerged from the cross-case analysis as common to all three countries. Within the visual conceptual framework (which depicts the interconnected nature of leading, teaching and learning in middle-level learning environments), these seven core categories were grouped into Essential Factors (student agency, voice and choice), Contextual Factors (instructional practices; school culture; engaging families and the community), Synergistic Factors (instructional leadership) and Cornerstone Factors (education as a fundamental cultural value; preservice, in-service and ongoing teacher development). In addition, sub-factors emerged from recurring codes in the data and identified specific characteristics and actions found in developmentally responsive, intellectually engaging middle-level learning environments. Although this study focused on 12 schools in Finland, Germany and Canada, it informs the practice of educators working with early adolescent learners in middle-level learning environments internationally. The authentic voices of early adolescent learners are the most important resource educators have to gauge if they are creating effective learning environments for their students. Ongoing professional dialogue and learning is essential to ensure teachers are supported in their work and develop the pedagogical practices needed to meet the needs of early adolescent learners. It is critical to balance consistency, coherence and dependability in the school environment with the necessary flexibility in order to support the unique learning needs of early adolescents. Educators must intentionally create a school culture that unites teachers, students and their families in support of a common purpose, as well as nurture positive relationships between the school and its community. A large, urban school district in Canada has implemented a school cohort-based model to begin to bring developmentally responsive, intellectually engaging middle-level learning environments to scale.

Keywords: Developmentally responsive learning environments, early adolescents, middle-level learning, middle years, instructional leadership, instructional practices, intellectually engaging learning environments, leadership dimensions, student agency.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1486

Authors: M. Othman, N. Bhuiyan, G. J. Gouw

Abstract:

In today-s competitive market, most companies develop manufacturing systems that can help in cost reduction and maximum quality. Human issues are an important part of manufacturing systems, yet most companies ignore their effects on production performance. This paper aims to developing an integrated workforce planning system that incorporates the human being. Therefore, a multi-objective mixed integer nonlinear programming model is developed to determine the amount of hiring, firing, training, overtime for each worker type. This paper considers a workforce planning model including human aspects such as skills, training, workers- personalities, capacity, motivation, and learning rates. This model helps to minimize the hiring, firing, training and overtime costs, and maximize the workers- performance. The results indicate that the workers- differences should be considered in workforce scheduling to generate realistic plans with minimum costs. This paper also investigates the effects of human learning rates on the performance of the production systems.

Keywords: Human Factors, Learning Curves, Workers' Differences, Workforce Scheduling

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1858

Authors: Miao Sabrina Wang

Abstract:

This paper examines the impact of grammatical differences on simultaneous interpreting from English into Mandarin Chinese by drawing upon an empirical study of professional and student interpreters. The research focuses on the effects of three grammatical categories including passives, adverbial components and noun phrases on simultaneous interpreting. For each category, interpretations of instances in which the grammatical structures are the same across the two languages are compared with interpretations of instances in which the grammatical structures differ across the two languages in terms of content accuracy and delivery appropriateness. The results indicate that grammatical differences have a significant impact on the interpreting performance of both professionals and students.

Keywords: Grammatical differences, simultaneous interpreting, content accuracy, delivery appropriateness.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1720

Authors: Muhammed Yousoof, Mohd Sapiyan, Khaja Kamaluddin

Abstract:

Computer programming is considered a very difficult course by many computer science students. The reasons for the difficulties include cognitive load involved in programming, different learning styles of students, instructional methodology and the choice of the programming languages. To reduce the difficulties the following have been tried: pair programming, program visualization, different learning styles etc. However, these efforts have produced limited success. This paper reviews the problem and proposes a framework to help students overcome the difficulties involved.

Keywords: Cognitive Load, Instructional Models, LearningStyles, Program Visualization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1446

Authors: Yi-Fan Zhu, Wei Zhang, Xuan Zhou, Qun Li, Yong-Lin Lei

Abstract:

The quick training algorithms and accurate solution procedure for incremental learning aim at improving the efficiency of training of SVR, whereas there are some disadvantages for them, i.e. the nonconvergence of the formers for changeable training set and the inefficiency of the latter for a massive dataset. In order to handle the problems, a new training algorithm for a changeable training set, named Approximation Incremental Training Algorithm (AITA), was proposed. This paper explored the reason of nonconvergence theoretically and discussed the realization of AITA, and finally demonstrated the benefits of AITA both on precision and efficiency.

Keywords: support vector regression, incremental learning, changeable training set, quick training algorithm, accurate solutionprocedure

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1476

Authors: Kripesh Adhikari, Hamid Bouchachia, Hammadi Nait-Charif

Abstract:

Falls are one of the major causes of injury and death among elderly people aged 65 and above. A support system to identify such kind of abnormal activities have become extremely important with the increase in ageing population. Pose estimation is a challenging task and to add more to this, it is even more challenging when pose estimations are performed on challenging poses that may occur during fall. Location of the body provides a clue where the person is at the time of fall. This paper presents a vision-based tracking strategy where available joints are grouped into three different feature points depending upon the section they are located in the body. The three feature points derived from different joints combinations represents the upper region or head region, mid-region or torso and lower region or leg region. Tracking is always challenging when a motion is involved. Hence the idea is to locate the regions in the body in every frame and consider it as the tracking strategy. Grouping these joints can be beneficial to achieve a stable region for tracking. The location of the body parts provides a crucial information to distinguish normal activities from falls.

Keywords: Fall detection, machine learning, deep learning, pose estimation, tracking.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2120

Authors: Abd Mutalib Embong, Azelin M Noor, Razol Mahari M Ali, Zulqarnain Abu Bakar, Abdur- Rahman Mohamed Amin

Abstract:

At the time where electronic books, or e-Books, offer students a fun way of learning , teachers who are used to the paper text books may find it as a new challenge to use it as a part of learning process. Precisely, there are various types of e-Books available to suit students- knowledge, characteristics, abilities, and interests. The paper discusses teachers- perceptions on the use of ebooks as a paper text book in the classroom. A survey was conducted on 72 teachers who use e-books as textbooks. It was discovered that a majority of these teachers had good perceptions on the use of ebooks. However, they had little problems using the devices. It can be overcome with some strategies and a suggested framework.

Keywords: Classroom, E-books, perception, teacher.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5743

Authors: Abobaker Mohammed Qasem Farhan, Shangming Yang, Mohammed Al-Nehari

Abstract:

Pneumonia is a kind of lung disease that creates congestion in the chest. Such pneumonic conditions lead to loss of life due to the severity of high congestion. Pneumonic lung disease is caused by viral pneumonia, bacterial pneumonia, or COVID-19 induced pneumonia. The early prediction and classification of such lung diseases help reduce the mortality rate. We propose the automatic Computer-Aided Diagnosis (CAD) system in this paper using the deep learning approach. The proposed CAD system takes input from raw computerized tomography (CT) scans of the patient's chest and automatically predicts disease classification. We designed the Hybrid Deep Learning Algorithm (HDLA) to improve accuracy and reduce processing requirements. The raw CT scans are pre-processed first to enhance their quality for further analysis. We then applied a hybrid model that consists of automatic feature extraction and classification. We propose the robust 2D Convolutional Neural Network (CNN) model to extract the automatic features from the pre-processed CT image. This CNN model assures feature learning with extremely effective 1D feature extraction for each input CT image. The outcome of the 2D CNN model is then normalized using the Min-Max technique. The second step of the proposed hybrid model is related to training and classification using different classifiers. The simulation outcomes using the publicly available dataset prove the robustness and efficiency of the proposed model compared to state-of-art algorithms.

Keywords: CT scans, COVID-19, deep learning, image processing, pneumonia, lung disease.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 588

Authors: Tonislav Ivanov, Oleksii Nedashkivskyi, Denis Babeshko, Vadim Pinskiy, Matthew Putman

Abstract:

Training a machine learning model for object detection that generalizes well is known to benefit from a training dataset with diverse examples. However, training datasets usually contain many repeats of common examples of a class and lack rarely seen examples. This is due to the process commonly used during human annotation where a person would proceed sequentially through a list of images labeling a sufficiently high total number of examples. Instead, the method presented involves an active process where, after the initial labeling of several images is completed, the next subset of images for labeling is selected by an algorithm. This process of algorithmic image selection and manual labeling continues in an iterative fashion. The algorithm used for the image selection is a deep learning algorithm, based on the U-shaped architecture, which quantifies the presence of unseen data in each image in order to find images that contain the most novel examples. Moreover, the location of the unseen data in each image is highlighted, aiding the labeler in spotting these examples. Experiments performed using semiconductor wafer data show that labeling a subset of the data, curated by this algorithm, resulted in a model with a better performance than a model produced from sequentially labeling the same amount of data. Also, similar performance is achieved compared to a model trained on exhaustive labeling of the whole dataset. Overall, the proposed approach results in a dataset that has a diverse set of examples per class as well as more balanced classes, which proves beneficial when training a deep learning model.

Keywords: Computer vision, deep learning, object detection, semiconductor.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 815

Authors: Ferdi Sönmez, Başak Buluz

Abstract:

Agility in Knowledge Management (AKM) tries to capture agility requirements and their respective answers within the framework of knowledge and learning for organizations. Since it is rather a new construct, it is difficult to claim that it has been sufficiently discussed and analyzed in practical and theoretical realms. Like the term ‘agile learning’, it is also commonly addressed in the software development and information technology fields and across the related areas where those technologies can be applied. The organizational perspective towards AKM, seems to need some more time to become scholarly mature. Nevertheless, in the literature one can come across some implicit usages of this term occasionally. This research is aimed to explore the conceptual background of agility in KM, re-conceptualize it and extend it to business applications with a special focus on e-business.

Keywords: Knowledge management, agility requirements, agility in knowledge management, knowledge.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1251

Authors: Kimberly Tao

Abstract:

This paper looks at transgender identities and the law in the context of marriage. It particularly focuses on the role of language and definition in classifying transgendered individuals into a legal category. Two lines of cases in transgender jurisprudence are examined. The former cases decided the definition of 'man' and 'woman' on the basis of biological criteria while the latter cases held that biological factors should not be the sole criterion for defining a man or a woman. Three categories were found to classify transgender people, namely male, female and "monstrous". Since transgender people challenge the core gender distinction that the law stresses, they are often regarded as problematic and monstrous which caused them to be subjected to severe legal consequences. This paper discusses these issues by analyzing and comparing different cases in transgender jurisprudence as well as examining how these issues play out in contemporary Hong Kong.

Keywords: Trangender, Monstrousness, Categorization, Definition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2188

Authors: Refaat M Mohamed, Ayman El-Baz, Aly A. Farag

Abstract:

This paper presents a new approach for the prob-ability density function estimation using the Support Vector Ma-chines (SVM) and the Expectation Maximization (EM) algorithms.In the proposed approach, an advanced algorithm for the SVM den-sity estimation which incorporates the Mean Field theory in the learning process is used. Instead of using ad-hoc values for the para-meters of the kernel function which is used by the SVM algorithm,the proposed approach uses the EM algorithm for an automatic optimization of the kernel. Experimental evaluation using simulated data set shows encouraging results.

Keywords: Density Estimation, SVM, Learning Algorithms, Parameters Estimation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2498

Authors: Dasimah Bt Omar, Faizul Abdullah, Fatimah Yusof, Hazlina Hamdan, Naasah Nasrudin, Ishak Che Abullah

Abstract:

The impacts of near-campus student housing, or offcampus students accommodation cannot be ignored by the universities and as well as the community officials. Numerous scholarly studies, have highlighted the substantial economic impacts either; direct, indirect or induced, and cumulatively the roles of the universities have significantly contributed to the local economies. The issue of the impacts of off-campus student rental housing on neighbourhoods is one that has been of long-standing but increasing concern in Malaysia. Statistically, in Malaysia, there was approximately a total of 1.2 - 1.5 million students in 2009. By the year 2015, it is expected that 50 per cent of 18 to 30 year olds active population should gain access to university education, amounting to 120,000 yearly. The objectives of the research are to assess the impacts off-campus students on the local neighbourhood and specifically to obtain information on the living and learning conditions of off-campus students of Universiti Teknologi MARA Shah Alam, Malaysia. It is also to isolate those factors that may impede the successful learning so that priority can be given to them in subsequent policy implementations and actions by government and the higher education institutions.

Keywords: off-campus students, neighbourhood, impacts, living and learning conditions

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 4403

Authors: Denise Levy, Anna Lucia C. H. Villavicencio

Abstract:

In a rapid-changing world, science teachers face considerable challenges. In addition to the basic curriculum, there must be included several transversal themes, which demand creative and innovative strategies to be arranged and integrated to traditional disciplines. In Brazil, nuclear science is still a controversial theme, and teachers themselves seem to be unaware of the issue, most often perpetuating prejudice, errors and misconceptions. This article presents the authors’ experience in the development of an interdisciplinary pedagogical proposal to include nuclear science in the basic curriculum, in a transversal and integrating way. The methodology applied was based on the analysis of several normative documents that define the requirements of essential learning, competences and skills of basic education for all schools in Brazil. The didactic materials and resources were developed according to the best practices to improve learning processes privileging constructivist educational techniques, with emphasis on active learning process, collaborative learning and learning through research. The material consists of an illustrated book for students, a book for teachers and a manual with activities that can articulate nuclear science to different disciplines: Portuguese, mathematics, science, art, English, history and geography. The content counts on high scientific rigor and articulate nuclear technology with topics of interest to society in the most diverse spheres, such as food supply, public health, food safety and foreign trade. Moreover, this pedagogical proposal takes advantage of the potential value of digital technologies, implementing QR codes that excite and challenge students of all ages, improving interaction and engagement. The expected results include the education of the educators for nuclear science communication in a transversal and integrating way, demystifying nuclear technology in a contextualized and significant approach. It is expected that the interdisciplinary pedagogical proposal contributes to improving attitudes towards knowledge construction, privileging reconstructive questioning, fostering a culture of systematic curiosity and encouraging critical thinking skills.

Keywords: Science education, interdisciplinary learning, nuclear science; scientific literacy.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 807

Authors: Barıs Kocer, Ahmet Arslan

Abstract:

Travelling salesman problem (TSP) is a combinational optimization problem and solution approaches have been applied many real world problems. Pure TSP assumes the cities to visit are fixed in time and thus solutions are created to find shortest path according to these point. But some of the points are canceled to visit in time. If the problem is not time crucial it is not important to determine new routing plan but if the points are changing rapidly and time is necessary do decide a new route plan a new approach should be applied in such cases. We developed a route plan transfer method based on transfer learning and we achieved high performance against determining a new model from scratch in every change.

Keywords: genetic algorithms, transfer learning, travellingsalesman problem

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1265

Authors: Juha Kettunen

Abstract:

The purpose of this study is to develop a framework for the assessment of research and development projects. The assessment map is developed in this study based on the strategy map of the balanced scorecard approach. The assessment map is applied in a project that aims to reduce the inequality and risk of exclusion of young people from disadvantaged social groups. The assessment map denotes that not only funding but also necessary skills and qualifications should be carefully assessed in the implementation of the project plans so as to achieve the objectives of projects and the desired impact. The results of this study are useful for those who want to develop the implementation of the Erasmus+ Programme and the project teams of research and development projects.

Keywords: Non-formal learning, youth work, social inclusion, innovation.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 823

Authors: Amina Abubakar

Abstract:

English like any other language is rich by means of arbitrary, conventional, symbols which lend it to lot of inconsistencies in spelling, phonology, syntax, and morphology. The research examines the irregularities prevalent in the structure and meaning of some ‘er’ lexical items in English and its implication to vocabulary acquisition. It centers its investigation on the derivational suffix ‘er’, which changes the grammatical category of word. English language poses many challenges to Second Language Learners because of its irregularities, exceptions, and rules. One of the meaning of –er derivational suffix is someone or somebody who does something. This rule often confuses the learners when they meet with the exceptions in normal discourse. The need to investigate instances of such inconsistencies in the formation of –er words and the meanings given to such words by the students motivated this study. For this purpose, some senior secondary two (SS2) students in six randomly selected schools in the metropolis were provided a large number of alphabetically selected ‘er’ suffix ending words, The researcher opts for a test technique, which requires them to provide the meaning of the selected words with- er. The marking of the test was scored on the scale of 1-0, where correct formation of –er word and meaning is scored one while wrong formation and meaning is scored zero. The number of wrong and correct formations of –er words meaning were calculated using percentage. The result of this research shows that a large number of students made wrong generalization of the meaning of the selected -er ending words. This shows how enormous the inconsistencies are in English language and how are affect the learning of English. Findings from the study revealed that though students mastered the basic morphological rules but the errors are generally committed on those vocabulary items that are not frequently in use. The study arrives at this conclusion from the survey of their textbook and their spoken activities. Therefore, the researcher recommends that there should be effective reappraisal of language teaching through implementation of the designed curriculum to reflect on modern strategies of teaching language, identification, and incorporation of the exceptions in rigorous communicative activities in language teaching, language course books and tutorials, training and retraining of teachers on the strategies that conform to the new pedagogy.

Keywords: ESL, derivational morpheme, inflectional morpheme, suffixes.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1893

Authors: Ava Dodd, Jonathan Adams

Abstract:

The purpose of this study is to explore the characteristics of developing a machine learning application using synthetic data. The study is structured to develop the application for the purpose of deploying the computer vision model. The findings discuss the realities of attempting to develop a computer vision model for practical purpose, and detail the processes, tools and techniques that were used to meet accuracy requirements. The research reveals that synthetic data represent another variable that can be adjusted to improve the performance of a computer vision model. Further, a suite of tools and tuning recommendations are provided.

Keywords: computer vision, machine learning, synthetic data, YOLOv4

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 839

Authors: Malathi Balakrishnan, Shabeshan Rengasamy, Mohd Salleh Aman

Abstract:

The study investigated the effects of Teaching Games for Understanding approach on students ‘cognitive learning outcome. The study was a quasi-experimental non-equivalent pretest-posttest control group design whereby 10 year old primary school students (n=72) were randomly assigned to an experimental and a control group. The experimental group students were exposed with TGfU approach and the control group with the Traditional Skill approach of handball game. Game Performance Assessment Instrument (GPAI) was used to measure students' tactical understanding and decision making in 3 versus 3 handball game situations. Analysis of covariance (ANCOVA) was used to analyze the data. The results reveal that there was a significant difference between the TGfU approach group and the traditional skill approach group students on post test score (F (1, 69) = 248.83, p < .05). The findings of this study suggested the importance of TGfU approach to improve primary students’ tactical understanding and decision making in handball game.

Keywords: Constructivism, learning outcome, tactical understanding, and Teaching Game for Understanding (TGfU)

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 4592

Authors: Eric Jiang

Abstract:

As a popular rank-reduced vector space approach, Latent Semantic Indexing (LSI) has been used in information retrieval and other applications. In this paper, an LSI-based content vector model for text classification is presented, which constructs multiple augmented category LSI spaces and classifies text by their content. The model integrates the class discriminative information from the training data and is equipped with several pertinent feature selection and text classification algorithms. The proposed classifier has been applied to email classification and its experiments on a benchmark spam testing corpus (PU1) have shown that the approach represents a competitive alternative to other email classifiers based on the well-known SVM and naïve Bayes algorithms.

Keywords: Feature Selection, Latent Semantic Indexing, Text Classification, Vector Space Model.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1882

Authors: Wei Zhang

Abstract:

With the rapid development of deep learning, neural network and deep learning algorithms play a significant role in various practical applications. Due to the high accuracy and good performance, Convolutional Neural Networks (CNNs) especially have become a research hot spot in the past few years. However, the size of the networks becomes increasingly large scale due to the demands of the practical applications, which poses a significant challenge to construct a high-performance implementation of deep learning neural networks. Meanwhile, many of these application scenarios also have strict requirements on the performance and low-power consumption of hardware devices. Therefore, it is particularly critical to choose a moderate computing platform for hardware acceleration of CNNs. This article aimed to survey the recent advance in Field Programmable Gate Array (FPGA)-based acceleration of CNNs. Various designs and implementations of the accelerator based on FPGA under different devices and network models are overviewed, and the versions of Graphic Processing Units (GPUs), Application Specific Integrated Circuits (ASICs) and Digital Signal Processors (DSPs) are compared to present our own critical analysis and comments. Finally, we give a discussion on different perspectives of these acceleration and optimization methods on FPGA platforms to further explore the opportunities and challenges for future research. More helpfully, we give a prospect for future development of the FPGA-based accelerator.

Keywords: Deep learning, field programmable gate array, FPGA, hardware acceleration, convolutional neural networks, CNN.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 884

Authors: Sokkhey Phauk, Takeo Okazaki

Abstract:

The challenging task in educational institutions is to maximize the high performance of students and minimize the failure rate of poor-performing students. An effective method to leverage this task is to know student learning patterns with highly influencing factors and get an early prediction of student learning outcomes at the timely stage for setting up policies for improvement. Educational data mining (EDM) is an emerging disciplinary field of data mining, statistics, and machine learning concerned with extracting useful knowledge and information for the sake of improvement and development in the education environment. The study is of this work is to propose techniques in EDM and integrate it into a web-based system for predicting poor-performing students. A comparative study of prediction models is conducted. Subsequently, high performing models are developed to get higher performance. The hybrid random forest (Hybrid RF) produces the most successful classification. For the context of intervention and improving the learning outcomes, a feature selection method MICHI, which is the combination of mutual information (MI) and chi-square (CHI) algorithms based on the ranked feature scores, is introduced to select a dominant feature set that improves the performance of prediction and uses the obtained dominant set as information for intervention. By using the proposed techniques of EDM, an academic performance prediction system (APPS) is subsequently developed for educational stockholders to get an early prediction of student learning outcomes for timely intervention. Experimental outcomes and evaluation surveys report the effectiveness and usefulness of the developed system. The system is used to help educational stakeholders and related individuals for intervening and improving student performance.

Keywords: Academic performance prediction system, prediction model, educational data mining, dominant factors, feature selection methods, student performance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 961

Authors: Salma El Hajjami, Jamal Malki, Alain Bouju, Mohammed Berrada

Abstract:

With the expansion of machine learning and data mining in the context of Big Data analytics, the common problem that affects data is class imbalance. It refers to an imbalanced distribution of instances belonging to each class. This problem is present in many real world applications such as fraud detection, network intrusion detection, medical diagnostics, etc. In these cases, data instances labeled negatively are significantly more numerous than the instances labeled positively. When this difference is too large, the learning system may face difficulty when tackling this problem, since it is initially designed to work in relatively balanced class distribution scenarios. Another important problem, which usually accompanies these imbalanced data, is the overlapping instances between the two classes. It is commonly referred to as noise or overlapping data. In this article, we propose an approach called: One Side Behavioral Noise Reduction (OSBNR). This approach presents a way to deal with the problem of class imbalance in the presence of a high noise level. OSBNR is based on two steps. Firstly, a cluster analysis is applied to groups similar instances from the minority class into several behavior clusters. Secondly, we select and eliminate the instances of the majority class, considered as behavioral noise, which overlap with behavior clusters of the minority class. The results of experiments carried out on a representative public dataset confirm that the proposed approach is efficient for the treatment of class imbalances in the presence of noise.

Keywords: Machine learning, Imbalanced data, Data mining, Big data.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1127

Authors: Saeed Mohammed Baneamoon, Rosalina Abdul Salam

Abstract:

Robotic system is an important area in artificial intelligence that aims at developing the performance techniques of the robot and making it more efficient and more effective in choosing its correct behavior. In this paper the distributed learning classifier system is used for designing a simulated control system for robot to perform complex behaviors. A set of enhanced approaches that support default hierarchies formation is suggested and compared with each other in order to make the simulated robot more effective in mapping the input to the correct output behavior.

Keywords: Learning Classifier System, Default Hierarchies, Robot Behaviors.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1419

Authors: Sandesh Achar

Abstract:

Artificial intelligence (AI) is being increasingly incorporated into many applications across various sectors such as health, education, security, and agriculture. Recently, there has been rapid development in cloud computing technology, resulting in AI’s implementation into cloud computing to enhance and optimize the technology service rendered. The deployment of AI in cloud-based applications has brought about autonomous computing, whereby systems achieve stated results without human intervention. Despite the amount of research into autonomous computing, work incorporating AI/ML into cloud computing to enhance its performance and resource allocation remains a fundamental challenge. This paper highlights different manifestations, roles, trends, and challenges related to AI-based cloud computing models. This work reviews and highlights investigations and progress in the domain. Future directions are suggested for leveraging AI/ML in next-generation computing for emerging computing paradigms such as cloud environments. Adopting AI-based algorithms and techniques to increase operational efficiency, cost savings, automation, reducing energy consumption and solving complex cloud computing issues are the major findings outlined in this paper.

Keywords: Artificial intelligence, AI, cloud computing, deep learning, machine learning, ML, internet of things, IoT.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 610

Authors: Salsabil Trabelsi, Zied Elouedi, Khaled Mellouli

Abstract:

The belief decision tree (BDT) approach is a decision tree in an uncertain environment where the uncertainty is represented through the Transferable Belief Model (TBM), one interpretation of the belief function theory. The uncertainty can appear either in the actual class of training objects or attribute values of objects to classify. In this paper, we develop a post-pruning method of belief decision trees in order to reduce size and improve classification accuracy on unseen cases. The pruning of decision tree has a considerable intention in the areas of machine learning.

Keywords: machine learning, uncertainty, belief function theory, belief decision tree, pruning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1905

Authors: Jianwei Wu

Abstract:

Based on the one-bit-matching principle and by turning the de-mixing matrix into an orthogonal matrix via certain normalization, Ma et al proposed a one-bit-matching learning algorithm on the Stiefel manifold for independent component analysis [8]. But this algorithm is not adaptive. In this paper, an algorithm which can extract kurtosis and its sign of each independent source component directly from observation data is firstly introduced.With the algorithm , the one-bit-matching learning algorithm is revised, so that it can make the blind separation on the Stiefel manifold implemented completely in the adaptive mode in the framework of natural gradient.

Keywords: Independent component analysis, kurtosis, Stiefel manifold, super-gaussians or sub-gaussians.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1497

Authors: Zilong He, Yuesheng Zhu

Abstract:

Based on the combined shape feature and texture feature, a fast object detection method with rotation invariant features is proposed in this paper. A quick template matching scheme based online learning designed for online applications is also introduced in this paper. The experimental results have shown that the proposed approach has the features of lower computation complexity and higher detection rate, while keeping almost the same performance compared to the HOG-based method, and can be more suitable for run time applications.

Keywords: gradient feature, online learning, rotationinvariance, template feature

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2471