Search Results for
"machine learning AND neural"

Recently, a strict working definition of information in computer science was proposed. Information is understood as a selected subset in relation to the original set of elements (alternatives, states, outcomes, etc.); in the general case, the selected subset can be fuzzy. The purpose of this study is to create a methodological basis for using the new definition of information in IT education, in particular when studying machine learning technologies and neural networks. In addition, an important methodological issue is considered: the relationship of the new definition of information with another central concept of information theory – diversity. The experience of implementing the new approach is analyzed using the example of a research and educational student project. The project is devoted to forecasting the emergence of evolutionarily stable daily vertical migrations of zooplankton as a result of adaptation to environmental conditions. In this case, external conditions are classified into four classes corresponding to different migration regimes of two age groups of zooplankton. The new approach to the definition of information involves taking into account the objective fuzziness of the boundaries between classes. This in turn leads to the fact that the forecast regarding the migrations is formed with varying degrees of confidence. This approach allows us to avoid unfounded decisions and reduce the number of errors. It significantly reduces the risks of neural network overtraining. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

This study focuses on developing a predictive modeling system to identify early signs of underperformance in vocational education, critical for building a skilled workforce. Addressing challenges like high dropout rates and inadequate graduate preparedness, the system utilizes machine learning techniques such as neural Networks, Decision Trees, and Logistic Regression. Implemented in Python, it analyzes key features like academic records, attendance, engagement, and socioeconomic factors. Preprocessing steps, such as data cleaning and feature engineering, were implemented, and transfer learning was employed to adapt the model. This combination of feature engineering and transfer learning enables the transfer of knowledge from academic settings to vocational education by identifying and leveraging shared characteristics between the two domains. The system provides real time insights through automated reports and notifications, enabling targeted interventions to improve retention and graduation rates. This scalable approach advances educational technology and informs policies to enhance vocational education outcomes. © 2025 IEEE.

In today's competitive job market, coupled with rapid technological developments and global events like the pandemic, securing employment has become increasingly difficult for new graduates. Therefore, this study aims to identify the key features that influence the employment status (employed or unemployed) and type (regular or nonregular) of university graduates in South Korea using several machine learning methods including logistic regression, decision tree, random forest, XGBoost, support vector machine, neural network, and Naïve Bayes. Among these methods, XGBoost and neural network demonstrated the highest performance. We applied the SHAP (SHapley Additive exPlanations), one of the XAI (eXplainable AI) techniques, to the XGBoost model. The results revealed that cumulative GPA is the most significant factor influencing labor market outcomes. Furthermore, geographical location, including the location of the university, high school, and current city of residence, as well as family background factors such as parents' occupation and income at the time of university admission, and current assets also played crucial roles. Overall, these results demonstrate the significant impact of academic performance, geographical location, and family background in shaping graduates' employment outcomes. These findings offer valuable insights for policymakers, educational institutions, students, and their families to develop strategies and policies for improving employment prospects. © 2025 KIIE

Vocational education systems worldwide are under increasing pressure to adapt their curricula to the demands of digital transformation, particularly in enhancing English language integration within vocational competency frameworks. However, most existing curriculum designs are static and fail to dynamically respond to evolving learner needs and industry standards. To address this challenge, this study proposes a deep learning-driven framework for optimizing the English-Vocational competency matrix using advanced machine learning models. A simulation-based dataset was generated to replicate learner performance scenarios, and three models, Decision Tree (DT), Support Vector machine (SVM), and neural Network (NN), were developed and tested using Python in Google Colab. Results showed that the neural Network significantly outperformed other models with an accuracy of 99.0%, precision of 100%, and F1-score of 99.1%, indicating its superior capability in capturing nonlinear patterns in educational data. This approach contributes a scalable and intelligent solution to dynamic curriculum adaptation in vocational education. The study concludes by recommending the integration of such models into adaptive learning platforms for real-time curriculum adjustment. The practical implication lies in enabling data-informed policy and pedagogical interventions, although limitations such as reliance on simulated data warrant future application to real-world datasets for validation. © 2025 IEEE.

In today's world, increasing stress and depression significantly impact cognitive well-being, making early detection of cognitive impairment essential for timely intervention. This work introduces a Multimodal Fusion Cognitive Assessment Framework that leverages advanced deep learning and graph intelligence to enhance early identification accuracy. Traditional tools like the Montreal Cognitive Assessment (MOCA) are limited in adaptability, prompting the need for a more dynamic, data-driven approach. The framework is validated using datasets involving cognitive tests, voice samples, and physiological signals. It enables a scalable, personalized, and adaptive cognitive assessment system that improves early detection and supports targeted intervention strategies. By integrating deep learning and information fusion, this approach addresses the complexity of cognitive health in a modern context. • This paper introduces Multimodal Deep learning Integration, incorporating MOCA scores, behavioral data, speech signals, and physiological parameters using GAT, TAT, and CNN-LSTM models to capture diverse cognitive indicators. • The proposed model achieves superior performance through Information Fusion via Heterogeneous GNNs, effectively merging cross-domain data to enable holistic cognitive state assessment via inter-modality learning. • This paper applies Reinforcement learning (RL) to personalize user interactions based on real-time cognitive and stress cues, reducing cognitive overload and enhancing engagement. © 2025 The Author(s)

Ecological sustainable development model is a green ecological model and a prerequisite for economic sustainable development. Higher education is vocational education and vocational training for relevant personnel, aiming to cultivate more professional talents and skilled workers for the country and society, so as to achieve the goal of social development. This paper aims to use advanced technical means to construct an ecological sustainable development model for higher education. Based on relevant computer technology, this paper proposes a support vector machine (SVM) model and a BP (back propagation) artificial neural network algorithm to analyze and study the ecological sustainable development model. Through experimental analysis, it is known that in the survey of teachers, 199 teachers believe that resources should be effectively utilized in the network environment. This part of teachers accounts for the largest proportion, which is 25.51%, while the proportion of teachers with academic theoretical knowledge is the smallest, which is 7.06%. This shows that teachers should strengthen technical learning. In the process of distance education, practical operation is far more important than theoretical knowledge, and teachers’ skill training also affects the sustainable development of education. Finally, the calculation results show that the experimental method has certain significance for the analysis of ecological sustainable development of higher education. © The Author(s) 2025.

Emotion level of the students during their academic sessions has a significant effect on their performance and overall academic grades. The most pre-eminent way of evaluating emotion levels is by analysing the EEG signals obtained from the brain. This paper showcases the experimental study on obtaining the Electroencephalogram (EEG) signals from the students during their academic sessions and classifying it to the types of emotions using machine learning algorithms. This paper explores the method of how the data collection session is conducted and recorded. The data collected is then compiled and divided for the machine learning algorithms. Furthermore, this paper proposed a method of acquiring the emotion labels without prior inducing by using a standard normal distribution method. Finally, this paper also proposed a deep learning neural network model and machine learning models that can be used to determine the emotion level from the EEG signals. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

The proceedings contain 56 papers. The special focus in this conference is on learning Ideas. The topics include: Literary History in Digital Teaching and learning: The KoLidi-Project—Collaborative and Interactive Approaches for German Studies; learning and Performance Science for Digital Transformation; a Conceptual Model for Meeting the Needs of Adult Learners in Distance Education; adapt to Learners: Practitioner Levels and Practice Support Methodologies; implementation of a Signature Pedagogy in an Online Course for Music Teachers; neural Correlates of Creative Drawing: Relationship Between EEG Output and a Domain-Specific Creativity Scale; experiential learning in Digital Contexts―A Case Study; experiences on Creating Personal Study Plans with Chatbots; the Notebook to Reflect on the Meaning of Life: An Educational Proposal for the Guidance of Young Migrants; socio-affective Profiles in Virtual learning Environments: Using learning Analytics; effects of Game Elements on Performances in Digital learning Games; creating a learning Environment for the Fifth Industrial Revolution; The 2CG® Poetry machine―a Hybrid Approach to Human Capability Cultivation with Disruptive Artistic Impulses; promoting Flourishing in Hard Times: Theoretical Reflections on Ethics of Care in Distance learning; a Project-Based learning Experience Through a Double Interaction Between Virtuality and Reality; models and Methods of Online Team Teaching; the Use of Comics as a Teaching and learning Tool; creating a Powerful Employee Experience: Lessons from Product Management; peer-to-Peer learning at Google and Peloton: The Power of Internal Experts; promoting Social Inclusion in Vocational Training Students with Disabilities: An Experience of Museum Education; adaptive Scaffolding Toward Transdisciplinary Collaboration: Reflective Polyvocal Self-study.

Entrepreneurship education has become essential in recent years. This education system may not be unconnected with the global agitation for value creation, employability skills and job creation. Engaging in entrepreneurial training provides students with the skills needed to enhance their ability to create marketable and profitable solutions to emerging problems. To do this, many emerging entrepreneurs rely on technology to engage in entrepreneurship education. This study presents a machine learning technique to predict the adaptability level of students in online entrepreneurship education. The suitability of different algorithms like Random Forest, C5.0, CART and Artificial neural Network was examined using the Kaggle Educational dataset. The algorithms recorded a high accuracy rate and affirmed machine learning techniques' ability to forecast students' adaptation to online entrepreneurship training. The findings of this research contribute to the field of online entrepreneurship education by providing a reliable and efficient approach for predicting students' adaptability. The proposed modified ensemble machine learning model can assist educators and administrators in identifying students who may require additional support, tailoring instructional strategies, and designing targeted interventions to enhance their adaptability and overall learning experience in online entrepreneurship education. © 2023