Online Machine Learning Research Articles

Ensemble adaptive online machine learning in data stream: a case study in cyber intrusion detection system

Ensemble adaptive online machine learning in data stream: a case study in cyber intrusion detection system

Dealing with Drift of Adaptation Spaces in Learning-based Self-Adaptive Systems Using Lifelong Self-Adaptation

Recently, machine learning (ML) has become a popular approach to support self-adaptation. ML has been used to deal with several problems in self-adaptation, such as maintaining an up-to-date runtime model under uncertainty and scalable decision-making. Yet, exploiting ML comes with inherent challenges. In this article, we focus on a particularly important challenge for learning-based self-adaptive systems: drift in adaptation spaces. With adaptation space, we refer to the set of adaptation options a self-adaptive system can select from to adapt at a given time based on the estimated quality properties of the adaptation options. A drift of adaptation spaces originates from uncertainties, affecting the quality properties of the adaptation options. Such drift may imply that the quality of the system may deteriorate, eventually, no adaptation option may satisfy the initial set of adaptation goals, or adaptation options may emerge that allow enhancing the adaptation goals. In ML, such a shift corresponds to a novel class appearance, a type of concept drift in target data that common ML techniques have problems dealing with. To tackle this problem, we present a novel approach to self-adaptation that enhances learning-based self-adaptive systems with a lifelong ML layer. We refer to this approach as lifelong self-adaptation . The lifelong ML layer tracks the system and its environment, associates this knowledge with the current learning tasks, identifies new tasks based on differences, and updates the learning models of the self-adaptive system accordingly. A human stakeholder may be involved to support the learning process and adjust the learning and goal models. We present a general architecture for lifelong self-adaptation and apply it to the case of drift of adaptation spaces that affects the decision-making in self-adaptation. We validate the approach for a series of scenarios with a drift of adaptation spaces using the DeltaIoT exemplar.

Training future engineers: Integrating Computational Thinking and effective learning methodologies into education

AbstractThis article examines the effectiveness and interest generated among primary and secondary education students through activities aimed at developing Computational Thinking skills, in the context of the coronavirus disease 2019 pandemic. The shift to online or hybrid learning models posed a significant challenge for educators, particularly those lacking digital skills. The study sought to answer several research questions, including the impact of online versus in‐person teaching on preuniversity students and gender differences in Computer Science perception, and Computational Thinking skills performance. The study employed a four‐phase methodology, consisting of pre‐ and posttraining measurements of Computer Science perception and Computational Thinking skills development through specific activities delivered in‐person or online. The results indicate that in‐person training is more effective for developing Computational Thinking skills, particularly at the secondary education level. Furthermore, there is a need to focus on maintaining girls' interest in Computer Science during primary school, as interest levels tend to decline significantly in secondary school. These findings have significant implications for Engineering Education in the context of digital transformation and the increasing importance of Computational Thinking skills in various fields of engineering. This study highlights the importance of developing Computational Thinking skills among preuniversity students and the need for effective training methods to achieve this goal and underscore the significance of investing in Engineering Education to prepare the next generation of engineers for the rapidly changing digital landscape.

CoRe optimizer: an all-in-one solution for machine learning

The optimization algorithm and its hyperparameters can significantly affect the training speed and resulting model accuracy in machine learning (ML) applications. The wish list for an ideal optimizer includes fast and smooth convergence to low error, low computational demand, and general applicability. Our recently introduced continual resilient (CoRe) optimizer has shown superior performance compared to other state-of-the-art first-order gradient-based optimizers for training lifelong ML potentials. In this work we provide an extensive performance comparison of the CoRe optimizer and nine other optimization algorithms including the Adam optimizer and resilient backpropagation (RPROP) for diverse ML tasks. We analyze the influence of different hyperparameters and provide generally applicable values. The CoRe optimizer yields best or competitive performance in every investigated application, while only one hyperparameter needs to be changed depending on mini-batch or batch learning.

DDPG-MPCC: An Experience Driven Multipath Performance Oriented Congestion Control

We introduce a novel multipath data transport approach at the transport layer referred to as ‘Deep Deterministic Policy Gradient for Multipath Performance-oriented Congestion Control’ (DDPG-MPCC), which leverages deep reinforcement learning to enhance congestion management in multipath networks. Our method combines DDPG with online convex optimization to optimize fairness and performance in simultaneously challenging multipath internet congestion control scenarios. Through experiments by developing kernel implementation, we show how DDPG-MPCC performs compared to the state-of-the-art solutions.

Improve the Quality of Online Teaching and Learning at Hung Vuong University on Technology Platform

In recent years, in order to innovate teaching methods and apply information technology to teaching, Hung Vuong University has applied an online teaching and learning model. However, this model has only been widely deployed to all sectors in the school when the nCov-19 epidemic broke out. This is a teaching method commonly applied in universities today. However, to be able to maximize the efficiency and quality of this model is not easy. This article proposes some solutions to improve the quality of online teaching and learning at Hung Vuong University.

The Online Saddle Point Problem and Online Convex Optimization with Knapsacks

We study the online saddle point problem, an online learning problem where at each iteration, a pair of actions needs to be chosen without knowledge of the current and future (convex-concave) payoff functions. The objective is to minimize the gap between the cumulative payoffs and the saddle point value of the aggregate payoff function, which we measure using a metric called saddle point regret (SP-Regret). The problem generalizes the online convex optimization framework, but here, we must ensure that both players incur cumulative payoffs close to that of the Nash equilibrium of the sum of the games. We propose an algorithm that achieves SP-Regret proportional to [Formula: see text] in the general case, and [Formula: see text] SP-Regret for the strongly convex-concave case. We also consider the special case where the payoff functions are bilinear and the decision sets are the probability simplex. In this setting, we are able to design algorithms that reduce the bounds on SP-Regret from a linear dependence in the dimension of the problem to a logarithmic one. We also study the problem under bandit feedback and provide an algorithm that achieves sublinear SP-Regret. We then consider an online convex optimization with knapsacks problem motivated by a wide variety of applications, such as dynamic pricing, auctions, and crowdsourcing. We relate this problem to the online saddle point problem and establish [Formula: see text] regret using a primal-dual algorithm.

Model pembelajaran online fisika berbasis pemodelan

Education in the discipline of physics has experienced significant changes along with developments in technology and internet access. One of these innovative approaches that is increasingly popular is modeling-based online learning models. This article aims to conduct a comprehensive literature review on modeling-based physics learning models implemented online. This approach emphasizes the use of modeling as one of the main tools to facilitate students' deeper understanding of physics concepts. In order to understand and evaluate its impact, this research adopts a structured approach by combining in-depth research and thorough analysis of this online learning model. The theory underlying this learning model involves the integration of information technology and modeling approaches in physics teaching. Supporting data and evidence were obtained through critical research on the implementation of this model in the context of online physics education. In addition, this research measures its effectiveness using data on students' understanding of physics concepts, student satisfaction questionnaires, and participatory observation. The results of this research indicate that the modeling-based online learning model is effective in increasing students' understanding of physics concepts. In-depth analysis also revealed that this approach provides a more interactive and engaging learning experience for students, which is reflected in high levels of satisfaction. This research confirms the great potential of modeling-based physics online learning models in addressing modern educational challenges. The practical and theoretical implications of these findings pave the way for further developments in online physics education and provide a basis for further research in this area. By integrating modeling in online physics learning, we can create a more dynamic and effective learning experience for students in the digital era.

AF-OS-ELM-MVE: A new online sequential extreme learning machine of dam safety monitoring model for structure deformation estimation

Dam deformation monitoring is an important task in hydraulic engineering projects. Traditional statistical methods and machine learning algorithms have been widely applied in structural safety prediction due to the application’s convenience. However, most machine learning algorithms are offline models and are not suitable for simulating the complex dynamic process of dam displacement. In addition, dam safety monitoring should quantify the uncertainty of the time-dependent displacement and establish reliable prediction intervals. As a result, an online learning model that can qualify prediction intervals is necessary. In this paper, an online learning algorithm named AF-OS-ELM-MVE is proposed to solve the problems. It is based on the OS-ELM algorithm and optimized using the adaptive forgetting factor mechanism. Through selective forgetting for parameter tuning, it achieves self-adaptation to the underlying physical mechanism between the factors and the displacement data. Mean Variance Estimation (MVE) is also adapted to establish the new hybrid model. It can be used to estimate the variance of the prediction process, and the model can generate prediction intervals based on the minimum probability theory. The AF-OS-ELM-MVE model is verified using long-term monitoring data and the performance comparisons were made with a number of state-of-the-art models. The correlation coefficients of the proposed model in point prediction are above 0.99 for measuring points of different elevations, and at a 95% confidence level, it achieves 100% coverage of real displacements. The results demonstrate the effectiveness of the proposed model in predicting dam deformations and provide a more solid basis for dam safety analysis.

A structural equation model of online learning: investigating self-efficacy, informal digital learning, self-regulated learning, and course satisfaction.

English as a Foreign Language (EFL) education increasingly relies on online learning, necessitating a nuanced understanding of crucial factors impacting learning experiences. This research investigates the intricate relationships among online learning self-efficacy, online self-regulated learning, informal digital learning of English (IDLE), and online course satisfaction within the unique context of EFL learners. The study involved 563 intermediate college students from various national universities in China. Structural Equation Modeling (SEM) was employed to analyze the data, providing comprehensive insights into the relationships among the identified variables. The results revealed significant insights. Both online learning self-efficacy and IDLE exhibited direct and positive influences on online course satisfaction. Furthermore, the study uncovered that online self-regulated learning acted as a partial mediator in the connection between online learning self-efficacy and IDLE with online course satisfaction. This mediation implies that learners' self-regulatory behaviors significantly affect how self-efficacy and informal digital language learning experiences impact overall satisfaction with online courses. The findings highlight the pivotal role of nurturing learners' self-efficacy beliefs, fostering IDLE, and promoting effective self-regulated learning strategies in the realm of online language learning. These initiatives are instrumental in enhancing learners' satisfaction and success in online courses. The implications of these findings for EFL instruction are substantial. By emphasizing the importance of self-efficacy, IDLE, and self-regulated learning strategies, educators can significantly contribute to creating more satisfying and successful online learning experiences for EFL students.

Building an Online Learning Model Through a Dance Recognition Video Based on Deep Learning

Jumping motion recognition via video is a significant contribution because it considerably impacts intelligent applications and will be widely adopted in life. This method can be used to train future dancers using innovative technology. Challenging poses will be repeated and improved over time, reducing the strain on the instructor when performing multiple times. Dancers can also be recreated by removing features from their images. To recognize the dancers’ moves, check and correct their poses, and another important aspect is that our model can extract cognitive features for efficient evaluation and classification, and deep learning is currently one of the best ways to do this for short-form video features capabilities. In addition, evaluating the quality of the performance video, the accuracy of each dance step is a complex problem when the eyes of the judges cannot focus 100% on the dance on the stage. Moreover, dance on videos is of great interest to scientists today, as technology is increasingly developing and becoming useful to replace human beings. Based on actual conditions and needs in Vietnam. In this paper, we propose a method to replace manual evaluation, and our approach is used to evaluate dance through short videos. In addition, we conduct dance analysis through short-form videos, thereby applying techniques such as deep learning to assess and collect data from which to draw accurate conclusions. Experiments show that our assessment is relatively accurate when the accuracy and F1-score values are calculated. More than 92.38% accuracy and 91.18% F1-score, respectively. This demonstrates that our method performs well and accurately in dance evaluation analysis.

The role of lifelong machine learning in bridging the gap between human and machine learning: A scientometric analysis

AbstractDue to advancements in data collection, storage, and processing techniques, machine learning has become a thriving and dominant paradigm. However, one of its main shortcomings is that the classical machine learning paradigm acts in isolation without utilizing the knowledge gained through learning from related tasks in the past. To circumvent this, the concept of Lifelong Machine Learning (LML) has been proposed, with the goal of mimicking how humans learn and acquire cognition. Human learning research has revealed that the brain connects previously learned information while learning new information from a single or small number of examples. Similarly, an LML system continually learns by storing and applying acquired information. Starting with an analysis of how the human brain learns, this paper shows that the LML framework shares a functional structure with the brain when it comes to solving new problems using previously learned information. It also provides a description of the LML framework, emphasizing its similarities to human brain learning. It also provides citation graph generation and scientometric analysis algorithms for the LML literatures, including information about the datasets and evaluation metrics that have been used in the empirical evaluation of LML systems. Finally, it presents outstanding issues and possible future research directions in the field of LML.This article is categorized under: Technologies > Machine Learning

Impact of Technological Advances on Educational Development: Reflections on Online Learning Models

Technological advances have led to many educational innovations that have had an impact on science, technology, engineering, and mathematics. In the face of a natural disaster such as C-19, it is clear that countries and governments can learn something from the response, as the combination of technology and education has enabled the world to weather difficult times. This paper will discuss the basic questions of technology and education. Including the impact of technology on education, methods for achieving the effective integration of education, and technology and methods for scientific and technological progress. Based on this, this paper presents future challenges reflected in online learning during the pandemic. For example, the conditions of electronic devices and networks required for online learning are not fully met by the versions of devices that all families can have, the reduction of teacher-student interaction, and student health issues. In addition, this paper will make appropriate solutions to this, corresponding to the previous problems, respectively: creating online and offline knowledge interoperability, utilizing web-based applications to strengthen the teacher-student connection, encouraging students to exercise at home with programs, and proposing reasonable solutions.

Co-Construction Online Learning Models in Early Childhood Education Management Courses

The challenge of online learning in higher education is that there is often a lack of focus on students taking part in lectures, which causes a loss of learning. This study aims to analyze the feasibility and effectiveness of the co-construction online learning model in early childhood education management courses. The research method uses a research and development approach with the ADDIE model. Data collection was carried out by interviews, questionnaires, and observations on 88 students taking early childhood education management courses. Product feasibility testing was carried out from questionnaire data analysis on 10 statement items using a four-rating scale, while observations and interviews were used to analyze the effectiveness of implementing the co-construction online learning model. Data analysis used qualitative descriptive and quantitative techniques to analyze the feasibility and effectiveness of implementing the co-construction online learning model. The results of the study in the electronic feasibility test for teaching materials achieved a score of 71.4%, which means it is feasible and does not need revision. While the effectiveness test was carried out twice with different timeframes, in the first data collection there were 73% of active students, then it increased in the second data collection with 88% of active students. This increase is to Skinner's theory of operant conditioning learning, where individual behavior is influenced by the repetition of actions received.

Model for Online Teaching and Learning in Zimbabwe Higher Education

This research sought to create a model for online teaching and learning in Zimbabwe higher education, having realised the haphazard manner in which online teaching and learning was being conducted in the said institutions. This was done through determination of the requirements for this endeavour and then using such to design a model. Comprehensive review of related literature was conducted, which resulted in the creation of a conceptual framework. The research followed the pragmatism research philosophy, backed by a mixed method research choice in which questionnaires and interviews were used as research instruments. A case study research design was adopted, with a population of 22000 students and 2240 lecturers. These were sampled at 1% of students and 5% of lecturers. Cluster sampling was used to split the groups, while stratified sampling was used in each group. The major findings of this study were that both lecturers and students require laptops and smartphones as key gadgets for OTL, the internet was also identified as a key infrastructural requirement for OTL success, ICT technical skills and online pedagogical skills are two key skills required especially by lecturers for them to be effective on the online space, and the challenges for OTL were mainly lack of gadgets, slow internet, power challenges, and skills gap. Key recommendations were for universities to come up with schemes to assist lecturers and students to acquire gadgets, continuous ICT skills upgrading for lecturers, online pedagogical skills training and provision of mitigatory measures to handle challenges to online teaching and learning.

Stochastic online convex optimization. Application to probabilistic time series forecasting

Stochastic online convex optimization. Application to probabilistic time series forecasting

Distributed Newton Step Projection Algorithm for Online Convex Optimization Over Time-Varying Unbalanced Networks

Distributed Newton Step Projection Algorithm for Online Convex Optimization Over Time-Varying Unbalanced Networks

Multi-label Lifelong Machine Learning: A Scoping Review of Algorithms, Techniques, and Applications

Multi-label Lifelong Machine Learning: A Scoping Review of Algorithms, Techniques, and Applications

Over-deposition assessment of Direct Energy Deposition (DED) using melt pool geometric features and Machine Learning

Over-deposition’s influence on the DED process is significant in terms of process stability and final part quality. Most recent studies on quality assessment are based on online monitoring of the melt pool and comparing it with pre-specified thresholds. In this study an online machine learning model is developed, capable of identifying two over-deposition defect types, satellite and wire-oscillation, by using frames captured by an optical camera targeting the melt pool. The features extracted from the images included five geometrical melt pool features and the operating laser power and were successfully used to achieve a 94.4% classification accuracy regardless of the final part geometry.

STUDENT’S PERCEPTIONS OF ONLINE AND HYBRID LEARNING IN INDONESIAN LANGUAGE COURSES

In the educational sector, the presence of the corona virus has had a positive impact. This is because this virous has been able to change the system of education in Indonesia, which initially only uses face to face learning model in the classroom, starting to get to know the online learning model and the hybrid learning model. The aim of this research is to describe the students’ perceptions of the implementation between hybrid learning at Majapahit Islamic University and the implementation of the online learning model at Universitas Terbuka (UT). This reserarch is qualitative research with a descriptive approach. The researchers uses survey techniques as data collection. The researchers use Miles and Huberman's interactive analysis technique as data analyses, namely (1) data collection, (2) data reduction, (3) data verification, and (4) data display. The research results show that the students from Universitas Terbuka give positive response to online learning via e-learning, while the students from Universitas Islam Majapahit also give a positive response to hybrid learning.