Search results for: unsupervised learning approach

Authors: A. Degale Desta, Cheng Jian

Abstract:

Deep learning application in computer vision is rapidly advancing, giving it the ability to monitor the public and quickly identify potentially anomalous behaviour from crowd scenes. Therefore, the purpose of the current work is to improve the performance of safety of people in crowd events from panic behaviour through introducing the innovative idea of Aggregation of Ensembles (AOE), which makes use of the pre-trained ConvNets and a pool of classifiers to find anomalies in video data with packed scenes. According to the theory of algorithms that applied K-means, KNN, CNN, SVD, and Faster-CNN, YOLOv5 architectures learn different levels of semantic representation from crowd videos; the proposed approach leverages an ensemble of various fine-tuned convolutional neural networks (CNN), allowing for the extraction of enriched feature sets. In addition to the above algorithms, a long short-term memory neural network to forecast future feature values and a handmade feature that takes into consideration the peculiarities of the crowd to understand human behavior. On well-known datasets of panic situations, experiments are run to assess the effectiveness and precision of the suggested method. Results reveal that, compared to state-of-the-art methodologies, the system produces better and more promising results in terms of accuracy and processing speed.

Keywords: action recognition, computer vision, crowd detecting and tracking, deep learning

Procedia PDF Downloads 151

Authors: Ryoo Hye Jin Agnes

Abstract:

This study aims to investigate the efficacy of clickers in fostering L2 class interaction. In an L2 classroom, active learner-to-learner interactions and learner-to-teacher interactions play an important role in language acquisition. In light of this, introducing learning tools that promote such interactions would benefit L2 classroom by fostering interaction. This is because the anonymity of clickers allows learners to express their needs without the social risks associated with speaking up in the class. clickers therefore efficiently help learners express their level of understanding during the process of learning itself. This allows for an evaluative feedback loop where both learners and teachers understand the level of progress of the learners, better enabling classrooms to adapt to the learners’ needs. Eventually this tool promotes participation from learners. This, in turn, is believed to be effective in fostering classroom interaction, allowing learning to take place in a more comfortable yet vibrant way. This study is finalized by presenting the result of an experiment conducted to verify the effectiveness of this approach when teaching pragmatic aspect of Korean expressions with similar semantic functions. The learning achievement of learners in the experimental group was found higher than the learners’ in a control group. A survey was distributed to the learners, questioning them regarding the efficacy of clickers, and how it contributed to their learning in areas such as motivation, self-assessment, increasing participation, as well as giving feedback to teachers. Analyzing the data collected from the questionnaire given to the learners, the study presented data suggesting that this approach increased the scope of interactivity in the classroom, thus not only increasing participation but enhancing the type of classroom participation among learners. This participation in turn led to a marked improvement in their communicative abilities.

Keywords: second language acquisition, interaction, clickers, learner response system, output from learners, learner’s cognitive process

Procedia PDF Downloads 513

Authors: Muhammad Bilal, Adil Ahmed

Abstract:

This study proposes an approach for electrical induction motor predictive maintenance utilizing machine learning algorithms. On the basis of a study of temperature data obtained from sensors put on the motor, the goal is to predict motor failures. The proposed models are trained to identify whether a motor is defective or not by utilizing machine learning algorithms like Support Vector Machines (SVM) and K-Nearest Neighbors (KNN). According to a thorough study of the literature, earlier research has used motor current signature analysis (MCSA) and vibration data to forecast motor failures. The temperature signal methodology, which has clear advantages over the conventional MCSA and vibration analysis methods in terms of cost-effectiveness, is the main subject of this research. The acquired results emphasize the applicability and effectiveness of the temperature-based predictive maintenance strategy by demonstrating the successful categorization of defective motors using the suggested machine learning models.

Keywords: predictive maintenance, electrical induction motors, machine learning, temperature signal methodology, motor failures

Procedia PDF Downloads 106

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 494

Authors: Nurul Nadrah

Abstract:

As a result of the COVID-19 pandemic, a movement control order was implemented, leading to the adoption of online learning as a substitute for conventional classroom instruction. Thus, this study aims to determine the differences in skill performance between online learning and conventional methods among nursing students. We employed a quasi-experimental design with purposive sampling, involving a total of 59 nursing students, and used online learning as the intervention. As a result, the study found there was a significant difference in student skill performance between online learning and conventional methods. As a conclusion, in times of hardship, it is necessary to implement alternative pedagogical approaches, especially in critical fields like nursing, to ensure the uninterrupted progression of educational programs. This study suggests that online learning can be effectively employed as a means of imparting knowledge to nursing students during their training.

Keywords: nursing education, online learning, skill performance, conventional learning method

Procedia PDF Downloads 32

Authors: Sumitra Nuanmeesri

Abstract:

The research focus on study, analyze and design the model of the infrastructure and computer networks for distance education, online learning via new media, e-learning and blended learning. The collected information from study and analyze process that information was evaluated by the index of item objective congruence (IOC) by 9 specialists to design model. The results of evaluate the model with the mean and standard deviation by the sample of 9 specialists value is 3.85. The results showed that the infrastructure and computer networks are designed to be appropriate to a great extent appropriate to a great extent.

Keywords: blended learning, new media, infrastructure and computer network, tele-education, online learning

Procedia PDF Downloads 397

Authors: Mohammad H. Taheri, David J. Brown, Nasser Sherkat

Abstract:

Engagement is one of the most important factors in determining successful outcomes and deep learning in students. Existing approaches to detect student engagement involve periodic human observations that are subject to inter-rater reliability. Our solution uses real-time multimodal multisensor data labeled by objective performance outcomes to infer the engagement of students. The study involves four students with a combined diagnosis of cerebral palsy and a learning disability who took part in a 3-month trial over 59 sessions. Multimodal multisensor data were collected while they participated in a continuous performance test. Eye gaze, electroencephalogram, body pose, and interaction data were used to create a model of student engagement through objective labeling from the continuous performance test outcomes. In order to achieve this, a type of continuous performance test is introduced, the Seek-X type. Nine features were extracted including high-level handpicked compound features. Using leave-one-out cross-validation, a series of different machine learning approaches were evaluated. Overall, the random forest classification approach achieved the best classification results. Using random forest, 93.3% classification for engagement and 42.9% accuracy for disengagement were achieved. We compared these results to outcomes from different models: AdaBoost, decision tree, k-Nearest Neighbor, naïve Bayes, neural network, and support vector machine. We showed that using a multisensor approach achieved higher accuracy than using features from any reduced set of sensors. We found that using high-level handpicked features can improve the classification accuracy in every sensor mode. Our approach is robust to both sensor fallout and occlusions. The single most important sensor feature to the classification of engagement and distraction was shown to be eye gaze. It has been shown that we can accurately predict the level of engagement of students with learning disabilities in a real-time approach that is not subject to inter-rater reliability, human observation or reliant on a single mode of sensor input. This will help teachers design interventions for a heterogeneous group of students, where teachers cannot possibly attend to each of their individual needs. Our approach can be used to identify those with the greatest learning challenges so that all students are supported to reach their full potential.

Keywords: affective computing in education, affect detection, continuous performance test, engagement, flow, HCI, interaction, learning disabilities, machine learning, multimodal, multisensor, physiological sensors, student engagement

Procedia PDF Downloads 83

Authors: Sagheer Ahmad

Abstract:

Teaching is a sacred profession whereby students are developed in their mental abilities to cope with the challenges of the remote world. Thinkers have developed plenty of interesting ways to make the learning process quick and absorbing for the students. However, third world countries are still lacking these remote facilities in the institutions, and therefore, teaching is totally dependent upon the skills of the teachers. Skillful teachers use self-devised and stimulating ideas to grab the attention of their students. Most of the time their ideas are based on local grounds with which the students are already familiar. This self-explanatory characteristic is the base of several local ideologies to disseminate scientific knowledge to new generations. Biology is such a subject which largely bases upon hypotheses, and teaching it in an interesting way is needful to create a friendly relationship between teacher and student, and to make a fantastic learning environment. Taxonomic classification if presented as it is, may not be attractive for the secondary school students who just start learning about biology at elementary levels. Presenting this hierarchy by exemplifying Kingdom, Phylum, Class, Order, family, genus and Species as comparatives of our division into continents, countries, cities, towns, villages, homes and finally individuals could be an attention-grabbing approach to make this concept get into bones of students. Similarly, many other interesting approaches have also been adopted to teach students in a fascinating way so that learning science subjects may not be boring for them. Discussing these appealing ways of teaching students can be a valuable stimulus to refine teaching methodologies about science, thereby promoting the concept of friendly learning.

Keywords: biology, innovative approaches, taxonomic classification, teaching

Procedia PDF Downloads 244

Authors: Szymon Rutkowski, Tomasz Korbak

Abstract:

We present a reinforcement-learning-based method of training neural machine translation models without parallel corpora. The standard encoder-decoder approach to machine translation suffers from two problems we aim to address. First, it needs parallel corpora, which are scarce, especially for low-resource languages. Second, it lacks psychological plausibility of learning procedure: learning a foreign language is about learning to communicate useful information, not merely learning to transduce from one language’s 'encoding' to another. We instead pose the problem of learning to translate as learning a policy in a communication game between two agents: the translator and the classifier. The classifier is trained beforehand on a natural language inference task (determining the entailment relation between a premise and a hypothesis) in the target language. The translator produces a sequence of actions that correspond to generating translations of both the hypothesis and premise, which are then passed to the classifier. The translator is rewarded for classifier’s performance on determining entailment between sentences translated by the translator to disciple’s native language. Translator’s performance thus reflects its ability to communicate useful information to the classifier. In effect, we train a machine translation model without the need for parallel corpora altogether. While similar reinforcement learning formulations for zero-shot translation were proposed before, there is a number of improvements we introduce. While prior research aimed at grounding the translation task in the physical world by evaluating agents on an image captioning task, we found that using a linguistic task is more sample-efficient. Natural language inference (also known as recognizing textual entailment) captures semantic properties of sentence pairs that are poorly correlated with semantic similarity, thus enforcing basic understanding of the role played by compositionality. It has been shown that models trained recognizing textual entailment produce high-quality general-purpose sentence embeddings transferrable to other tasks. We use stanford natural language inference (SNLI) dataset as well as its analogous datasets for French (XNLI) and Polish (CDSCorpus). Textual entailment corpora can be obtained relatively easily for any language, which makes our approach more extensible to low-resource languages than traditional approaches based on parallel corpora. We evaluated a number of reinforcement learning algorithms (including policy gradients and actor-critic) to solve the problem of translator’s policy optimization and found that our attempts yield some promising improvements over previous approaches to reinforcement-learning based zero-shot machine translation.

Keywords: agent-based language learning, low-resource translation, natural language inference, neural machine translation, reinforcement learning

Procedia PDF Downloads 124

Authors: Seyyed Hassan Seyyedrezaei

Abstract:

The advent of computer and ICT thereafter has introduced many irrevocable changes in learning and teaching. There is substantially growing need for the use of IT and ICT in language learning and teaching. In other words, the integration of Information Technology (IT) into online teaching is of vital importance for education and assessment. Considering the fact that the image of education is undergone drastic changes by the advent of technology, education systems and teachers move beyond the walls of traditional classes and methods in order to join with other educational centers to revitalize education. Given the advent of distance learning, online courses and virtual universities, e-assessment has taken a prominent place in effective teaching and meeting the learners' educational needs. The purpose of this paper is twofold: first, scrutinizing e-learning, it discusses how and why e-assessment is becoming widely used by educationalists and administrators worldwide. As a second purpose, a couple of effective strategies for online assessment will be enumerated.

Keywords: e-assessment, e learning, ICT, online assessment

Procedia PDF Downloads 562

Authors: Sudhir Kumar Singh, Debashish Chakravarty

Abstract:

Slope stability analysis is an important aspect in the field of geotechnical engineering. It is also important from safety, and economic point of view as any slope failure leads to loss of valuable lives and damage to property worth millions. This paper aims at mitigating the risk of slope failure by studying the effect of different joint sets on the stability of mine pit slopes under the influence of various external factors, namely degree of saturation, rainfall intensity, and seismic coefficients. Supervised machine learning approach has been utilized for making accurate and reliable predictions regarding the stability of slopes based on the value of Factor of Safety. Numerous cases have been studied for analyzing the stability of slopes using the popular Finite Element Method, and the data thus obtained has been used as training data for the supervised machine learning models. The input data has been trained on different supervised machine learning models, namely Random Forest, Decision Tree, Support vector Machine, and XGBoost. Distinct test data that is not present in training data has been used for measuring the performance and accuracy of different models. Although all models have performed well on the test dataset but Random Forest stands out from others due to its high accuracy of greater than 95%, thus helping us by providing a valuable tool at our disposition which is neither computationally expensive nor time consuming and in good accordance with the numerical analysis result.

Keywords: finite element method, geotechnical engineering, machine learning, slope stability

Procedia PDF Downloads 95

Authors: Kafi Razzaq Ahmed

Abstract:

Speaking fluently in a second language requires vocabulary, grammar, and pronunciation skills. Teaching the English language entails teaching pronunciation. In professional literature, there have been a lot of attempts to integrate technology into improving the pronunciation of learners. The technique is also neglected in Kurdish contexts, Salahaddin University – Erbil included. Thus, the main aim of the research is to point out the efficiency of using video materials for both language teachers and learners within and beyond classroom learning and teaching environments to enhance student's pronunciation. To collect practical data, a research project has been designed. In subsequent research, a posttest will be administered after each lesson to 100 first-year students at Salahaddin University-Erbil English departments. All students will be taught the same material using different methods, one based on video materials and the other based on the traditional approach to teaching pronunciation. Finally, the results of both tests will be analyzed (also knowing the attitudes of both the teachers and the students about both lessons) to indicate the impact of using video in the process of teaching and learning pronunciation.

Keywords: video, pronunciation, teaching, learning

Procedia PDF Downloads 102

Authors: Hani Mustafa, Cristina Gonzalez Ruiz, Estelle Bech

Abstract:

The growing influence of digital technologies has made learning and interaction more accessible, resulting in effective collaboration if properly managed. Technology enabled learning has become an important conduit for learning, including collaborative learning. The use of wiki technology, for example, has opened a new learning platform that enables the integration of social, linguistic, and cognitive processes of language learning. It encourages students to collaborate in the construction, analysis, and understanding of knowledge. But to what extent is the use of wikis effective in promoting collaborative learning among students. In addition, how do students perceive this technology in enhancing their language learning? In this study, students were be given a wiki project to complete collaboratively with their group members. Students had to write collaboratively to produce and present a seven-day travel plan in which they had to describe places to visit and things to do to explore the best historical and cultural aspects of the country. The study involves students learning French, Malay, and Spanish as a foreign language. In completing this wiki project, students will move from passive learning of language to real engagement with classmates, requiring them to collaborate and negotiate effectively with one another. The objective of the study is to ascertain to what extent does wiki technology helped in promoting collaborative learning and improving language skills from students’ perception. It is found that while there was improvement in students language skills, the overall experience was less positive due to unfamiliarity with a new learning tool.

Keywords: collaborative learning, foreign language, wiki, teaching

Procedia PDF Downloads 134

Authors: Vinai K. Singh

Abstract:

In Neural Network-based Learning techniques, there are several models of Convolutional Networks. Whenever the methods are deployed with large datasets, only then can their applicability and appropriateness be determined. Clinical and pathological pictures of lobular carcinoma are thought to exhibit a large number of random formations and textures. Working with such pictures is a difficult problem in machine learning. Focusing on wet laboratories and following the outcomes, numerous studies have been published with fresh commentaries in the investigation. In this research, we provide a framework that can operate effectively on raw photos of various resolutions while easing the issues caused by the existence of patterns and texturing. The suggested approach produces very good findings that may be used to make decisions in the diagnosis of cancer.

Keywords: lobular carcinoma, convolutional neural networks (CNN), deep learning, histopathological imagery scans

Procedia PDF Downloads 130

Authors: Ghodhbani Salah, Jemili Farah

Abstract:

With the growth of cyber attacks, information safety has become an important issue all over the world. Many firms rely on security technologies such as intrusion detection systems (IDSs) to manage information technology security risks. IDSs are considered to be the last line of defense to secure a network and play a very important role in detecting large number of attacks. However the main problem with today’s most popular commercial IDSs is generating high volume of alerts and huge number of false positives. This drawback has become the main motivation for many research papers in IDS area. Hence, in this paper we present a data mining technique to assist network administrators to analyze and reduce false positive alarms that are produced by an IDS and increase detection accuracy. Our data mining technique is unsupervised clustering method based on hybrid ANT algorithm. This algorithm discovers clusters of intruders’ behavior without prior knowledge of a possible number of classes, then we apply K-means algorithm to improve the convergence of the ANT clustering. Experimental results on real dataset show that our proposed approach is efficient with high detection rate and low false alarm rate.

Keywords: intrusion detection system, alarm filtering, ANT class, ant clustering, intruders’ behaviors, false alarms

Procedia PDF Downloads 400

Authors: Hyeongbok Kim, Lingling Zhao, Xiaohong Su

Abstract:

Deep reinforcement learning has been applied to address various problems in robotics, such as autonomous driving and unmanned aerial vehicle. However, because of the sparse reward penalty for a collision with obstacles during the navigation mission, the agent fails to learn the optimal policy or requires a long time for convergence. Therefore, using obstacles and enemy agents, in this paper, we present a curriculum-based boost learning method to effectively train compound skills during multi-agent reinforcement learning. First, to enable the agents to solve challenging tasks, we gradually increased learning difficulties by adjusting reward shaping instead of constructing different learning environments. Then, in a benchmark environment with static obstacles and moving enemy agents, the experimental results showed that the proposed curriculum learning strategy enhanced cooperative navigation and compound collision avoidance skills in uncertain environments while improving learning efficiency.

Keywords: curriculum learning, hard exploration, multi-agent reinforcement learning, robotic navigation, sparse reward

Procedia PDF Downloads 85

Authors: Kazeem Oluwatoyin Ajape

Abstract:

The study is concerned with the problem of how to improve the teaching of Arabic as a foreign language in Nigerian Higher Education System. The paper traces the historical background of Arabic education in Nigeria and also outlines the problems facing the language in Nigerian Institutions. It lays down some of the essential foundation work necessary for bringing about systematic and constructive improvements in the Teaching of Arabic as a Foreign Language (TAFL) by giving answers to the following research questions: what is the appropriate medium of instruction in teaching a foreign or second language? What is the position of English language in the teaching and learning of Arabic/Islamic education? What is the relevance of the present curriculum of Arabic /Islamic education in Nigerian institutions to the contemporary society? A survey of the literature indicates that a revolution is currently taking place in FL teaching and that a new approach known as the Communicative Approach (CA), has begun to emerge and influence the teaching of FLs in general, over the last decade or so. Since the CA is currently being adapted to the teaching of most major FLs and since this revolution has not yet had much impact on TAPL, the study explores the possibility of the application of the CA to the teaching of Arabic as a living language and also makes recommendations towards the development of the language in Nigerian Institutions of Higher Learning.

Keywords: Arabic Language, foreign language, Nigerian institutions, curriculum, communicative approach

Procedia PDF Downloads 599

Authors: Winnie Wing Mui So, Yu Chen

Abstract:

This study investigated primary school teachers’ and students’ perceptions of science learning in an e-learning environment. This study used a multiple case study design and involved eight science teachers and their students from four Hong Kong primary schools. The science topics taught included ‘season and weather’ ‘force and movement’, ‘solar and lunar eclipse’ and ‘living things and habitats’. Data were collected through lesson observations, interviews with teachers, and interviews with students. Results revealed some differences between the teachers’ and the students’ perceptions regarding the usefulness of e-learning resources, the organization of student-centred activities, and the impact on engagement and interactions in lessons. The findings have implications for the more effective creation of e-learning environments for science teaching and learning in primary schools.

Keywords: e-learning, science education, teacher' and students' perceptions, primary schools

Procedia PDF Downloads 195

Authors: Irfan Ahmed Rind

Abstract:

Science education has recently become the top priority of government of Pakistan. Number of schemes has been initiated for the improvement of science teaching and learning at primary and secondary levels of education, most importantly training in-service science teachers on inquiry based teaching and learning to empower students and encourage creativity, critical thinking, and innovation among them. Therefore, this approach has been promoted in the recent continuous professional development trainings for the in-service teachers. However, the follow ups on trained science teachers and educators suggest that these teachers fail to implement the inquiry based teaching and learning in their classes. In addition, these trainings also fail to bring any significant change in students’ science content knowledge and understanding as per the annual national level surveys conducted by government and independent agencies. Research suggests that science has been taught using scientific positivism, which supports objectivity based on experiments and mathematics. In contrary, the inquiry based teaching and learning are based on constructivism, which conflicts with the positivist epistemology of science teachers. It was, therefore, assumed that science teachers struggle to implement the inquiry based teaching approach as it conflicts with their basic epistemological beliefs. With this assumption, this research aimed to (i) understand how science teachers conceptualize the nature of science, and how this influence their understanding of learning, learners, their own roles as teachers and their teaching strategies, (ii) identify the conflict of science teachers’ epistemological beliefs with the inquiry based teaching approach, and (iii) find the ways in which science teachers epistemological beliefs may be developed from positivism to constructivism, so that they may effectively use the inquiry based teaching approach in teaching science. Using qualitative case study approach, thirty six secondary and higher secondary science teachers (21 male and 15 female) were selected. Data was collected using interviewed, participatory observations (sixty lessons were observed), and twenty interviews from students for verifications of teachers’ responses. The findings suggest that most of the science teacher were positivist in defining the nature of science. Most of them limit themselves to one fix answer that is provided in the books and that there is only one 'right' way to teach science. There is no room for students’ or teachers’ own opinion or bias when it comes to scientific concepts. Inquiry based teaching seems 'no right' to them. They find it difficult to allow students to think out of the box. However, some interesting exercises were found to be very effective in bringing the change in teachers’ epistemological beliefs. These will be discussed in detail in the paper. The findings have major implications for the teachers, educators, and policymakers.

Keywords: science teachers, epistemology, metacognition, inquiry based teaching

Procedia PDF Downloads 145

Authors: Frida C. Jaime Franco, C. Paulina Navarro Núñez, R. Jacob Sánchez Nájera

Abstract:

The importance of developing multi-language abilities in our global society is noteworthy. However, the necessity, interest, and consciousness of the significance that the development of another language represents, apart from the mother tongue, is not always the same in all contexts as it is in multicultural communities, especially in rural higher education institutions immersed in small communities. Leading opportunities for digital interaction among learners from Mexico and abroad partners represents scaffolding towards, not only language skills development but also intercultural communicative competences (ICC). This study leads us to consider what should be the best approach to work while applying a program of ICC integrated into the practice of EFL. While analyzing the roots of the language, it is possible to obtain the main objective of learning another language, to communicate with a functional purpose, as well as attaching social practices to the learning process, giving a result of functionality and significance to the target language. Hence, the collateral impact that collaborative learning leads to, aims to contribute to a better global understanding as well as a means of self and other cultural awareness through intercultural communication. While communicating through the target language by online collaboration among students in platforms of long-distance communication, language is used as a tool of interaction to broaden students’ perspectives reaching a substantial improvement with the help of their differences. This process should consider the application of the target language in the inquiry of sociocultural information, expecting the learners to integrate communicative skills to handle cultural differentiation at the same time they apply the knowledge of their target language in a real scenario of communication, despite being through virtual resources.

Keywords: collaborative learning, communicative approach, culture, interaction, interculturalism, target language, virtual partnership

Procedia PDF Downloads 121

Authors: Wayne DeFehr

Abstract:

This study is rooted in a pedagogical approach that emphasizes student engagement as fundamental to meaningful learning in the classroom. This approach creates a paradigmatic shift, from a teaching practice that reinforces the teacher’s central authority to a practice that disperses that authority among the students in the classroom through networks that they themselves develop. The methodology of this study about creating optimal conditions for learning in the classroom includes providing a conceptual framework within which the students work, as well as providing clearly stated expectations for work standards, content quality, group methodology, and learning outcomes. These learning conditions are nurtured in a variety of ways. First, nearly every class includes a lecture from the professor with key concepts that students need in order to complete their work successfully. Secondly, students build on this scholarly material by forming their own networks, where students face each other and engage with each other in order to collaborate their way to solving a particular problem relating to the course content. Thirdly, students are given short, medium, and long-term goals. Short term goals relate to the week’s topic and involve workshopping particular issues relating to that stage of the course. The medium-term goals involve students submitting term assignments that are evaluated according to a well-defined rubric. And finally, long-term goals are achieved by creating a capstone project, which is celebrated and shared with classmates and interested friends on the final day of the course. The essential conclusions of the study are drawn from courses that focus on video game narrative. Enthusiastic student engagement is created not only with the dynamic energy and expertise of the instructor, but also with the inter-dependence of the students on each other to build knowledge, acquire skills, and achieve successful results.

Keywords: collaboration, education, learning networks, video games

Procedia PDF Downloads 111

Authors: Shayan Mohajer Hamidi

Abstract:

Federated learning has emerged as a promising approach for distributed model training while preserving data privacy. However, the challenges of communication overhead, limited network resources, and slow convergence hinder its widespread adoption. On the other hand, knowledge distillation has shown great potential in compressing large models into smaller ones without significant loss in performance. In this paper, we propose an innovative framework that combines federated learning and knowledge distillation to address these challenges and enhance the efficiency of distributed learning. Our approach, called Federated Knowledge Distillation (FKD), enables multiple clients in a federated learning setting to collaboratively distill knowledge from a teacher model. By leveraging the collaborative nature of federated learning, FKD aims to improve model compression while maintaining privacy. The proposed framework utilizes a coded teacher model that acts as a reference for distilling knowledge to the client models. To demonstrate the effectiveness of FKD, we conduct extensive experiments on various datasets and models. We compare FKD with baseline federated learning methods and standalone knowledge distillation techniques. The results show that FKD achieves superior model compression, faster convergence, and improved performance compared to traditional federated learning approaches. Furthermore, FKD effectively preserves privacy by ensuring that sensitive data remains on the client devices and only distilled knowledge is shared during the training process. In our experiments, we explore different knowledge transfer methods within the FKD framework, including Fine-Tuning (FT), FitNet, Correlation Congruence (CC), Similarity-Preserving (SP), and Relational Knowledge Distillation (RKD). We analyze the impact of these methods on model compression and convergence speed, shedding light on the trade-offs between size reduction and performance. Moreover, we address the challenges of communication efficiency and network resource utilization in federated learning by leveraging the knowledge distillation process. FKD reduces the amount of data transmitted across the network, minimizing communication overhead and improving resource utilization. This makes FKD particularly suitable for resource-constrained environments such as edge computing and IoT devices. The proposed FKD framework opens up new avenues for collaborative and privacy-preserving distributed learning. By combining the strengths of federated learning and knowledge distillation, it offers an efficient solution for model compression and convergence speed enhancement. Future research can explore further extensions and optimizations of FKD, as well as its applications in domains such as healthcare, finance, and smart cities, where privacy and distributed learning are of paramount importance.

Keywords: federated learning, knowledge distillation, knowledge transfer, deep learning

Procedia PDF Downloads 64

Authors: Kika Ioannou Kazamia

Abstract:

This paper examines the action research cycle of the second phase of longitudinal research on sustainable interior design practices, between two groups of stakeholders, designers and clients. During this phase of the action research, the second step - the change stage - of Lewin’s change management model has been utilized to change values, approaches, and attitudes toward sustainable design practices among the participants. Affective domain learning theory is utilized to attach new values. Learning with the use of information technology, collaborative learning, and problem-based learning are the learning methods implemented toward the acquisition of the objectives. Learning methods, and aims, require the design of interventions with participants' involvement in activities that would lead to the acknowledgment of the benefits of sustainable practices. Interventions are steered to measure participants’ decisions for the worth and relevance of ideas, and experiences; accept or commit to a particular stance or action. The data collection methods used in this action research are observers’ reports, participants' questionnaires, and interviews. The data analyses use both quantitative and qualitative methods. The main beneficial aspect of the quantitative method was to provide the means to separate many factors that obscured the main qualitative findings. The qualitative method allowed data to be categorized, to adapt the deductive approach, and then examine for commonalities that could reflect relevant categories or themes. The results from the data indicate that during the second phase, designers and clients' participants altered their behaviours.

Keywords: design, change, sustainability, learning, practices

Procedia PDF Downloads 72

Authors: Elena Puica

Abstract:

This paper aims to investigate the impact of Supply Chain Management (SCM) on the environment by applying a Machine Learning model while pointing out the efficiency of the technology used. The Machine Learning model was used to derive the efficiency and optimization of technology used in SCM and the environmental impact of SCM processes. The model applied is a predictive classification model and was trained firstly to determine which stage of the SCM has more outputs and secondly to demonstrate the efficiency of using advanced technology in SCM instead of recuring to traditional SCM. The outputs are the emissions generated in the environment, the consumption from different steps in the life cycle, the resulting pollutants/wastes emitted, and all the releases to air, land, and water. This manuscript presents an innovative approach to applying advanced technology in SCM and simultaneously studies the efficiency of technology and the SCM's impact on the environment. Identifying the conceptual relationships between SCM practices and their impact on the environment is a new contribution to the research. The authors can take a forward step in developing recent studies in SCM and its effects on the environment by applying technology.

Keywords: machine-learning model in SCM, SCM processes, SCM and the environmental impact, technology in SCM

Procedia PDF Downloads 110

Authors: Ilana Lavy, Rami Rashkovits

Abstract:

In this study, we describe Information Systems students' learning model which was applied by students in order to cope with professional knowledge gaps in the context of their final project. The students needed to implement a software system according to specifications and design they have made beforehand. They had to select certain technologies and use them. Most of them decided to use programming environments that were learned during their academic studies. The students had to cope with various levels of knowledge gaps. For that matter they used learning strategies that were organized by us as a learning model which includes two phases each suitable for different learning tasks. We analyze the learning model, describing advantages and shortcomings as perceived by the students, and provide excerpts to support our findings.

Keywords: knowledge gaps, independent learner skills, self-regulated learning, final project

Procedia PDF Downloads 474

Authors: Olubukola Olumuyiwa Tokede, Dominic D. Ahiaga-Dagbui, John Morrison

Abstract:

The project environment in construction has been widely criticised for its inability to learn from experience effectively. As each project is bespoke, learning is ephemeral, as it is often confined within its bounds and seldom assimilated with others that are being delivered in the project environment. To engender learning across construction projects, collaborative contractual arrangements, such as alliancing and partnering, have been embraced to aid the transferability of lessons across projects. These cooperative arrangements, however, tend to be costly, and hence construction organisations could revert to less expensive traditional procurement approaches after successful collaborative project delivery. This research, therefore, seeks to assess the lessons-learned in a post-alliance contractual framework. Using a case-study approach, we examine the experiences of a public sector authority who engaged a project facilitator to foster learning during the delivery of a significant piece of critical infrastructure. It was found that the facilitator enabled optimal learning outcomes in post-alliance contractual frameworks by attenuating the otherwise adversarial relationship between clients and contractors. Further research will seek to assess the effectiveness of different knowledge-brokering agencies in construction projects.

Keywords: facilitation, knowledge-brokering, learning, projects

Procedia PDF Downloads 132

Authors: S. A. N. Danushka, T. A. Weerasinghe

Abstract:

The diverse necessities of instruction could be addressed effectively with the support of new dimensions of ICT integrated learning such as blended learning –which is a combination of face-to-face and online instruction which ensures greater flexibility in student learning and congruity of course delivery. As blended learning has been the ‘new normality' in education, many experimental and quasi-experimental research studies provide ample of evidence on its successful implementation in many fields of studies, but it is hard to justify whether blended learning could work similarly in the delivery of technology-teacher development programmes (TTDPs). The present study is bound with the particular research uncertainty, and having considered existing research approaches, the study methodology was set to decide the efficient instructional strategies for flipped classroom learning in TTDPs. In a quasi-experimental pre-test and post-test design with a mix-method research approach, the major study objective was tested with two heterogeneous samples (N=135) identified in a virtual learning environment in a Sri Lankan university. Non-randomized informal ‘before-and-after without control group’ design was employed, and two data collection methods, identical pre-test and post-test and Likert-scale questionnaires were used in the study. Selected two instructional strategies, self-directed learning (SDL) and self-regulated learning (SRL), were tested in an appropriate instructional framework with two heterogeneous samples (pre-service and in-service teachers). Data were statistically analyzed, and an efficient instructional strategy was decided via t-test, ANOVA, ANCOVA. The effectiveness of the two instructional strategy implementation models was decided via multiple linear regression analysis. ANOVA (p < 0.05) shows that age, prior-educational qualifications, gender, and work-experiences do not impact on learning achievements of the two diverse groups of learners through the instructional strategy is changed. ANCOVA (p < 0.05) analysis shows that SDL is efficient for two diverse groups of technology-teachers than SRL. Multiple linear regression (p < 0.05) analysis shows that the staged self-directed learning (SSDL) model and four-phased model of motivated self-regulated learning (COPES Model) are efficient in the delivery of course content in flipped classroom learning.

Keywords: COPES model, flipped classroom learning, self-directed learning, self-regulated learning, SSDL model

Procedia PDF Downloads 188

Authors: Saba Rahimi, Ozan Oktay, Javier Alvarez-Valle, Sujeeth Bharadwaj

Abstract:

Successful application of deep learning in medical image analysis necessitates unprecedented amounts of labeled training data. Unlike conventional 2D applications, radiological images can be three-dimensional (e.g., CT, MRI), consisting of many instances within each image. The problem is exacerbated when expert annotations are required for effective pixel-wise labeling, which incurs exorbitant labeling effort and cost. Active learning is an established research domain that aims to reduce labeling workload by prioritizing a subset of informative unlabeled examples to annotate. Our contribution is a cost-effective approach for U-Net 3D models that uses Monte Carlo sampling to analyze pixel-wise uncertainty. Experiments on the AAPM 2017 lung CT segmentation challenge dataset show that our proposed framework can achieve promising segmentation results by using only 42% of the training data.

Keywords: image segmentation, active learning, convolutional neural network, 3D U-Net

Procedia PDF Downloads 146

Authors: Basiliyos Tilahun Betru, Charles Awono Onana, Bernabe Batchakui

Abstract:

Recommender systems have become increasingly popular in recent years, and are utilized in numerous areas. Nowadays many web services provide several information for users and recommender systems have been developed as critical element of these web applications to predict choice of preference and provide significant recommendations. With the help of the advantage of deep learning in modeling different types of data and due to the dynamic change of user preference, building a deep model can better understand users demand and further improve quality of recommendation. In this paper, deep neural network models for recommender system are evaluated. Most of deep neural network models in recommender system focus on the classical collaborative filtering user-item setting. Deep learning models demonstrated high level features of complex data can be learned instead of using metadata which can significantly improve accuracy of recommendation. Even though deep learning poses a great impact in various areas, applying the model to a recommender system have not been fully exploited and still a lot of improvements can be done both in collaborative and content-based approach while considering different contextual factors.

Keywords: big data, decision making, deep learning, recommender system

Procedia PDF Downloads 469

Authors: Jazim Sohail, Filipe Teixeira-Dias

Abstract:

Mild traumatic brain injury (mTBI) within non-helmeted contact sports is a growing concern due to the serious risk of potential injury. Extensive research is being conducted looking into head kinematics in non-helmeted contact sports utilizing instrumented mouthguards that allow researchers to record accelerations and velocities of the head during and after an impact. This does not, however, allow the location of the impact on the head, and its magnitude and orientation, to be determined. This research proposes and validates two methods to quantify impact locations from instrumented mouthguard kinematic data, one using rigid body dynamics, the other utilizing machine learning. The rigid body dynamics technique focuses on establishing and matching moments from Euler’s and torque equations in order to find the impact location on the head. The methodology is validated with impact data collected from a lab test with the dummy head fitted with an instrumented mouthguard. Additionally, a Hybrid III Dummy head finite element model was utilized to create synthetic kinematic data sets for impacts from varying locations to validate the impact location algorithm. The algorithm calculates accurate impact locations; however, it will require preprocessing of live data, which is currently being done by cross-referencing data timestamps to video footage. The machine learning technique focuses on eliminating the preprocessing aspect by establishing trends within time-series signals from instrumented mouthguards to determine the impact location on the head. An unsupervised learning technique is used to cluster together impacts within similar regions from an entire time-series signal. The kinematic signals established from mouthguards are converted to the frequency domain before using a clustering algorithm to cluster together similar signals within a time series that may span the length of a game. Impacts are clustered within predetermined location bins. The same Hybrid III Dummy finite element model is used to create impacts that closely replicate on-field impacts in order to create synthetic time-series datasets consisting of impacts in varying locations. These time-series data sets are used to validate the machine learning technique. The rigid body dynamics technique provides a good method to establish accurate impact location of impact signals that have already been labeled as true impacts and filtered out of the entire time series. However, the machine learning technique provides a method that can be implemented with long time series signal data but will provide impact location within predetermined regions on the head. Additionally, the machine learning technique can be used to eliminate false impacts captured by sensors saving additional time for data scientists using instrumented mouthguard kinematic data as validating true impacts with video footage would not be required.

Keywords: head impacts, impact location, instrumented mouthguard, machine learning, mTBI

Procedia PDF Downloads 210