Figure - uploaded by Victor Osasu Eguavoen
Content may be subject to copyright.
Source publication
This study introduces a novel hybrid soft-computing model integrating Fuzzy C-Means (FCM) clustering, Particle Swarm Optimization (PSO), and Adaptive Neuro-Fuzzy Inference System (ANFIS) to enhance student academic performance prediction in higher education. We developed a hybrid FCM-PSO-ANFIS model using a comprehensive dataset encompassing pre-ad...
Context in source publication
Context 1
... preprocessed dataset revealed a typical attrition pattern in longitudinal educational studies, with student numbers decreasing from 238 in the first year to 130 in the third and fourth years, reflecting factors like academic progression, program transfers, or withdrawals, which were considered during analysis. Table 1 presents the key variables used in the study, their descriptions, categorizations, and variable types. This categorization approach was designed to facilitate subsequent analysis and enhance the interpretability of results. ...Citations
... Education in the digital age has undergone significant transformations, with the integration of information and communication technologies (ICT) offering unprecedented opportunities for innovation in teaching and learning in rapidly evolving educational landscapes and increasing student diversity, the ability to accurately and proactively manage student academic performance has become more critical than ever (Eguavoen and Nwelih, 2024). Traditional educational systems, while effective in their era, often fail to address the diverse needs of modern learners, particularly in adapting to varying learning styles, paces, and preferences (Demir, 2021). ...
Smart education systems have emerged as pivotal tools in modern education, yet the effectiveness of different recommendation algorithms in personalizing learning experiences remains understudied, particularly in higher education contexts. This study evaluates and compares three recommendation algorithms collaborative filtering, content-based filtering, and a hybrid approach within a web-based smart education system designed for computer science education, focusing on their ability to enhance personalized learning experiences. The author developed a web-based smart education system incorporating these three algorithms and tested it using a dataset of 100 users across 20 courses. The system's performance was evaluated using precision, recall, accuracy, and F1-score metrics. A four-week case study with 20 users was conducted to assess practical implementation outcomes. The hybrid algorithm demonstrated superior performance with 85.96% precision, 98.99% recall, 94.33% accuracy, and a 91.98% F1-score, significantly outperforming both collaborative filtering (precision: 68.09%, recall: 94.12%) and content-based filtering (precision: 72.73%, recall: 90.32%). Case study results showed consistent improvement in user performance, with score improvements ranging from 15% to 23%.The hybrid algorithm proves most effective for personalizing educational content delivery, though with higher computational overhead. These findings suggest that hybrid recommendation approaches can significantly enhance smart education systems' ability to provide personalized learning experiences, despite computational challenges in large-scale deployments.
