Online Machine Learning Research Articles

3-D geometric design of microwaveable food products for optimal heating uniformity based on machine learning-supervised multiphysics models

Multiphysics models can assist in geometric design of frozen microwaveable foods, but the conventional 'parametric-sweeping' strategy is computationally intensive. This study develops an online machine learning (ML)-supervised multiphysics modeling strategy to simultaneously optimize multiple geometrical parameters (e.g., top surface width, top surface length, and the ratio of top-to-bottom dimensions) for optimal microwave heating uniformity. First, one or more paired geometric dimensional parameters and heating uniformity results obtained from multiphysics modeling are used as initial training data for the ML optimization process that integrates Gaussian Process Regression (GPR) and Bayesian optimization. Then, the multiphysics modeling procedure is supervised by an ML process to generate new paired geometry-heating uniformity results to expand the training dataset. The loop of multiphysics modeling and ML optimization is conducted to effectively identify geometry designs with good heating uniformity. Results indicate that the ML-supervised optimization strategy can efficiently identify good geometric designs with low heating uniformity by using much fewer multiphysics models than the parametric sweep approach. The ML-supervised approach also exhibits great robustness, delivering good performance even with small (as little as one model) and randomly selected initial training data. This ML-supervised approach is flexible and can be modified to meet specific needs for broad industrial implementation in the development of microwaveable food.

Open-Ended Online Learning for Autonomous Visual Perception.

The visual perception systems aim to autonomously collect consecutive visual data and perceive the relevant information online like human beings. In comparison with the classical static visual systems focusing on fixed tasks (e.g., face recognition for visual surveillance), the real-world visual systems (e.g., the robot visual system) often need to handle unpredicted tasks and dynamically changed environments, which need to imitate human-like intelligence with open-ended online learning ability. Therefore, we provide a comprehensive analysis of open-ended online learning problems for autonomous visual perception in this survey. Based on "what to online learn" among visual perception scenarios, we classify the open-ended online learning methods into five categories: instance incremental learning to handle data attributes changing, feature evolution learning for incremental and decremental features with the feature dimension changed dynamically, class incremental learning and task incremental learning aiming at online adding new coming classes/tasks, and parallel and distributed learning for large-scale data to reveal the computational and storage advantages. We discuss the characteristic of each method and introduce several representative works as well. Finally, we introduce some representative visual perception applications to show the enhanced performance when using various open-ended online learning models, followed by a discussion of several future directions.

Online Learning Models for Vehicle Usage Prediction During COVID-19

Online Learning Models for Vehicle Usage Prediction During COVID-19

Knowledge transfer in lifelong machine learning: a systematic literature review

Lifelong Machine Learning (LML) denotes a scenario involving multiple sequential tasks, each accompanied by its respective dataset, in order to solve specific learning problems. In this context, the focus of LML techniques is on utilizing already acquired knowledge to adapt to new tasks efficiently. Essentially, LML concerns about facing new tasks while exploiting the knowledge previously gathered from earlier tasks not only to help in adapting to new tasks but also to enrich the understanding of past ones. By understanding this concept, one can better grasp one of the major obstacles in LML, known as Knowledge Transfer (KT). This systematic literature review aims to explore state-of-the-art KT techniques within LML and assess the evaluation metrics and commonly utilized datasets in this field, thereby keeping the LML research community updated with the latest developments. From an initial pool of 417 articles from four distinguished databases, 30 were deemed highly pertinent for the information extraction phase. The analysis recognizes four primary KT techniques: Replay, Regularization, Parameter Isolation, and Hybrid. This study delves into the characteristics of these techniques across both neural network (NN) and non-neural network (non-NN) frameworks, highlighting their distinct advantages that have captured researchers’ interest. It was found that the majority of the studies focused on supervised learning within an NN modelling framework, particularly employing Parameter Isolation and Hybrid for KT. The paper concludes by pinpointing research opportunities, including investigating non-NN models for Replay and exploring applications outside of computer vision (CV).

Online learning MPC for switched systems with performance dependent mixed switching law

This paper explores online learning model predictive control (MPC) for switched systems with stable terminal constraints. In order to reduce the computational burden, we decompose the infinite horizon MPC into n finite horizons, and adopt adaptive dynamic programming (ADP) to assist MPC in online solving the optimal control problem in each finite horizon. Further, we use a set of stable terminal constraints to ensure both the convergence of online learning MPC in each finite horizon and the connection of adjacent finite horizons. In addition, in order to ensure the uniform ultimate boundedness (UUB) of the triggered switched systems, on the one hand, a novel performance dependent mixed switching law (PD-MSL) is proposed to both avoid frequent switching and take advantage of performance dependent decision-making; on the other hand, an analytical framework of the coupling between mode switching and event-triggering is proposed. Finally, the validity of the presented approach is demonstrated through simulations.

Addressing Data Distribution Shifts in Online Machine Learning Powered Smart City Applications Using Augmented Test-Time Adaptation

Addressing Data Distribution Shifts in Online Machine Learning Powered Smart City Applications Using Augmented Test-Time Adaptation

The Using of Distance Learning (Online) to Students of SMP IT Alhijrah Bintuju

This research was carried out at SMP IT Alhijrah Bintuju. The aim of this research is to find out whether there is a significant influence of the application of distance learning techniques (online) on the ability to write procedure texts by class VII students at SMP IT Alhijrah Bintuju.This research is an experimental study with a population of class VII SMP IT Alhijrah Bintuju consisting of 1 class in which 28 students. Meanwhile, the sample for this research was class VII with a total of 28 students using total sampling techniques. The data from this research were obtained from test scores, namely a test of the ability to write procedure texts before using distance learning techniques (pre test) and a test of the ability to write procedure texts after using distance learning techniques (post test). The data obtained were analyzed using the t test formula.Based on the results of data analysis, the results obtained consist of: 1) The results of data calculations illustrate that the mean score for the ability to write procedure texts after using the distance learning technique (online) was obtained with a score of 78.04, in the "Good" category. 2) The results of data calculations illustrate that the mean score for the ability to write procedure texts before using the distance learning (online) learning model was obtained with a score of 67.68, in the "Enough" category. 3) The results of hypothesis testing obtained a value of t0 = 4.42. These results are compared with tt at degrees of freedom (df) 54 or (N1 + N2 - 2) = 28 + 28 - 2 = 54. The tt score is 2.00 at the 5% significance level. Thus, it can be concluded that t0 is greater than tt, namely = 4.42 2.0. So it can be said "there is a significant influence of the use of distance (online) learning techniques on the ability to write procedure texts by class VII students at SMP IT Alhijrah Bintuju." Keywords: Distance Learning Techniques, Online, Procedure Text, Writing

Wireless Local Area Network and Analytical Hierarchy Process: Technology Acceptance Model for Increasing Use of Online Learning

Online learning is currently very much needed for the development of science in schools, the covid 19 outbreak has made changes to learning methods. The change in the learning model must also be supported by supporting learning devices and media, including computer networks and the internet in schools, the network that is used to support online learning is currently widely used by schools using WLAN, because it can reach parts of the hard one. The use of WLAN learning devices and media as a support for online learning can be an obstacle for some teachers and students, this is because not all teachers and students can understand the online learning system and not all areas have a fast internet access network, with these constraints in need a technology acceptance measurement model (TAM) for online learning using WLAN, so that online learning can be carried out properly and the achievement of successful learning can be carried out.

Distributed online constrained convex optimization with event-triggered communication

This paper focuses on the distributed online convex optimization problem with time-varying inequality constraints over a network of agents, where each agent collaborates with its neighboring agents to minimize the cumulative network-wide loss over time. To reduce communication overhead between the agents, we propose a distributed event-triggered online primal–dual algorithm over a time-varying directed graph. With several classes of appropriately chose decreasing parameter sequences and non-increasing event-triggered threshold sequences, we establish dynamic network regret and network cumulative constraint violation bounds. Finally, a numerical simulation example is provided to verify the theoretical results.

Learning models and achievements of the most gifted students

The emergency of switching to an online learning model due to the COVID pandemic impacted the educational process in of schools world wide. We conducted an anonymous survey at the elite high schools in Serbia whose students won 45 medals at international Olympiads in mathematics, physics, and computer sciences in the last 4-year period (2019–2022). The impact of switching from in-person to online learning on the achievement of the gifted students was analyzed. The results of the research revealed a downward trend in the final grades and achievements in international Olympiads, which persists even after returning to the face-to-face educational model. This trend is confirmed by calculation of the Pearson correlation coefficient (−0.62) between school years and achievements in international mathematical Olympiads, international physics Olympiads, and international Olympiads in informatics. Gifted students’ satisfaction with online learning during the COVID pandemic was also analyzed using content analysis of open-ended questions. The surveyed gifted students showed a high level of self-criticism and maturity by expressing their opinion about online learning. In the group of gifted students, there is a higher level of agreement about harmful consequences of online learning on the level of their education and work habits than in other research conducted on similar topics. Moreover, the level of agreement is in a positive correlation with students’ achievements and final grades.

Geological risk prediction under uncertainty in tunnel excavation using online learning and hidden Markov model

Geological risk prediction under uncertainty in tunnel excavation using online learning and hidden Markov model

Online learning and continuous model upgrading with data streams through the Kafka-ML framework

A pipeline of constant data streams is being built by the Internet of Things (IoT) to monitor information about the physical environment. In parallel, Artificial Intelligence (AI) is constantly developing and enhancing industrial, economic, and academic endeavors as well as quality of life thanks to these IoT data. In streaming contexts, Kafka-ML is our open-source framework that enables the management of Machine Learning (ML) and AI pipelines over data streams. Accordingly, it simplifies the deployment of Deep Neural Networks (DNNs) in practical applications. Nonetheless, this framework did not support the possibility of carrying out an Online Learning (OL) process, which is needed when new data are continuously arriving, and the models need to adapt to them on the fly. In this work, we have extended our previous work, the Kafka-ML framework, to enhance the management of ML/AI pipelines with OL features to enable both ML/AI distributed and centralized models to learn indefinitely over time. These models are continuously upgraded thanks to a process where automatic and flexible inference is carried out when improvements in the model performance are achieved. This opens up a large number of new possibilities within different fields of application, development, and work under the premise of incremental learning with ML models such as Electrical Vehicles and Industry 5.0. We have validated these new features by adapting and deploying state-of-the-art DNN models in different online scenarios, for both single and distributed configurations. The results show the capability of Kafka-ML to execute effective online training processes for ML models, improving their performance over time as new data becomes available.

The effectiveness of parental assistance-based online learning model on self-efficacy

<span lang="EN-US">The obstacles experienced during online learning are inevitable circumstances. Hence, students need to have motivational beliefs that they have the ability to thrive in online learning despite its difficulties. Applying parental-based online learning is expected to help students increase their self-efficacy. This study examines the effect of using an online learning model based on parental assistance on students’ self-efficacy with a quantitative approach. This type of research is quasi-experimental, with a pre and post-test control group design by studying seventh-grade students in several public junior high schools in Kisaran. The study found that applying an online learning model based on parental assistance improved students’ self-efficacy more than a conventional online learning model. Consequently, parental assistance will make a practical contribution to the implementation of learning, especially in the implementation of online learning in the era of the pandemic and after the COVID-19 pandemic.</span>

Assortment Optimization: An Annotated Reading Assortment

Which varieties or brands of a product should a retailer stock on its shelf? Carrying a large variety caters to more customers' needs, but could cannibalize the sales of high-end brands and also cause an inventory nightmare. Assortment optimization aims to formalize these tradeoffs, with the basic problem being as follows. There is a universe of brands j ∈ U , each with a market-accepted price r j . For any S ⊆ U , a function ϕ ( j, S ) indicates the probability that a representative customer from the population would purchase j when given the choice from assortment S , under the market prices. The optimization problem is to maximize the average revenue per customer, i.e. [EQUATION] (1) possibly with constraints on S due to shelf size. Assortment optimization started out by showing how to efficiently find the optimal S from the exponentially many possibilities, under well-established parametric forms for the function ϕ that are called ( discrete) choice models. Since then, the literature has developed choice models of its own that are specialized for assortment optimization. The basic problem has also been extended, and connected with topics such as online algorithms, machine learning, and mechanism design that are mainstream in the Economics and Computation community, with a vast horizon for future directions. This is an annotated reading list about assortment optimization, that aims to provide broad coverage while facing a "cardinality constraint" on the number of papers in the assortment.

Combining Regularization With Look-Ahead for Competitive Online Convex Optimization

Combining Regularization With Look-Ahead for Competitive Online Convex Optimization

Lightweight Distributed Gaussian Process Regression for Online Machine Learning

Lightweight Distributed Gaussian Process Regression for Online Machine Learning

EXPLORING TRENDS IN PUBLIC-PRIVATE PARTNERSHIPS IN VOCATIONAL EDUCATION: A FOCUS ON THE BALTIC COUNTRIES

This scientific article delves into the dynamic landscape of vocational education in the Baltic countries – Estonia, Latvia and Lithuania. It focuses on the evolving trends in public-private partnerships (PPPs) within this sector. Vocational education serves as a crucial link between academic learning and industry demands, equipping individuals with the specialized skills needed for the workforce. In recent years, significant strides have been made in enhancing vocational education in the Baltic region through collaborative efforts between government bodies, educational institutions and private enterprises. This article particularly emphasizes emerging trends in PPPs, including joint initiatives, government incentives, digitalization and the impact of the COVID-19 pandemic. Joint initiatives between educational institutions and industry stakeholders are becoming increasingly prevalent, fostering collaboration in curriculum development, internship programmes and skills training tailored to meet industry needs. Government incentives, such as financial support and tax benefits, are encouraging private sector involvement in vocational education, thereby enriching the quality and accessibility of educational offerings. Moreover, digitalization is reshaping vocational education, with PPPs leveraging technology to deliver innovative learning experiences. E-learning platforms, virtual reality simulations and online collaboration tools are being integrated into vocational education curricula, enhancing the effectiveness and relevance of training programmes. However, the COVID-19 pandemic has posed unprecedented challenges, accelerating the adoption of online learning and remote delivery models. By analyzing these trends, the article aims to provide insights into the dynamics of vocational education development and the role of collaboration between public and private sectors in shaping the future of the workforce in the Baltic countries.

Tutorial and Lecture Activities on Deep Online Learning in Dentistry

Online learning has become a new normal in the era of the COVID-19 pandemic. However, this learning mode has brought many challenges that can impact the learning experience. The study aims to evaluate student feedback on the online learning model and identify any challenges that require attention to enhance the quality of Dentistry education at Muhammadiyah University of Yogyakarta. This research is a descriptive observational study that utilized a cross-sectional approach. A modified questionnaire comprising 27 questions was distributed to collect the data. It was found that most of the 277 surveyed students agreed that implementing online learning in tutorial activities had a positive impact. Notably, 47.3% of the students found the online tutorial activities to be an excellent way to learn the material, while an equal number of students, 47.3% believed that they developed their tolerance during the online tutorial group sessions. Furthermore, the student responses to the online lecture activities were also positive, with 40.8% of the students finding comfort in the recorded classes, while 37.2% better understood the material through the online lectures. As for student satisfaction, the majority of the surveyed students expressed their satisfaction with the online learning experience regarding the material 52%, learning strategy 40,4%, availability of e-resources 49,1%, and assistance provided 42,6%. Students responded positively to online learning in tutorials and lectures and reported overall satisfaction with online learning during the COVID-19 pandemic.

ONLINE LEARNING MODEL TO IMPROVE THE STUDENTS’ ACHIEVEMENT IN DESIGN OF INFORMATION AND COMMUNICATION TECHNOLOGY

The rapid advancement of technology and the growing demand for ICT skills have necessitated innovative educational approaches to improve student achievement. Traditional classroom-based teaching often struggles to keep pace with the dynamic and evolving nature of ICT education. The variability in student learning styles and the increasing requirement for digital skills necessitate a flexible and scalable educational model. This study focuses on developing an effective online learning model tailored to enhance student performance in the field of Design of Information and Communication Technology (ICT). The study employed a mixed-methods approach, combining quantitative and qualitative data. Participants included undergraduate students enrolled in an ICT Design course. They were given access to an online learning platform offering interactive video lectures, multimedia resources, assessments, and collaborative projects. Data collection included pre- and post-tests to measure academic performance, surveys to gauge student satisfaction, and focus group discussions to gather qualitative insights into their learning experience. The findings indicated a significant improvement in student achievement, with post-test scores consistently higher than pre-test scores across all participants. The model's flexibility allowed students to learn at their own pace, leading to a deeper understanding of key concepts. Interactive resources and collaborative projects enhanced engagement and fostered critical thinking. Moreover, the majority of students expressed high satisfaction with the online learning experience, citing the variety of resources, self-paced learning opportunities, and collaborative forums as significant contributors.

Online Learning Model and Assistance of Parents and Private Teachers in Increasing Student Learning Achievement at MIMA KH. Shiddiq Jember

This article discusses the online learning model and the assistance of parents and private teachers in improving student learning achievement at MIMA KH. Siddiq Jember. Online learning is not only due to an epidemic, but the demands of contemporary educational developments which require using online. This research uses qualitative methods and data collection uses observation and interviews. The results of this research show that parents and private teachers play a very active role in improving student achievement because there is close cooperation between the school and parents or private teachers at home. So that it can build a learning environment that can encourage students to be active and really learn, and can also help educate students to catch up on lessons they have missed at school.