This study introduces a web-based application designed to facilitate hyperparameter optimization for machine learning models, leveraging data from the stack overflow dataset. A primary contribution of this research is the development of a novel hyperparameter optimization method, integrated alongside established techniques such as Grid Search, Random Search, Nelder-Mead, and Bayesian Optimization. This integration provides users with the flexibility to explore various optimization strategies and identify the most suitable approach for their specific datasets and models.
The proposed web application enables users to select datasets, define optimization methods, and fine-tune hyperparameters through an intuitive and user-friendly interface. Empirical results demonstrate that the optimized models achieved a 15% improvement in prediction accuracy, attaining approximately 70% accuracy in predicting coding expertise and programming languages. Furthermore, the Proposed Method enhanced memory efficiency by 20% in SVM-based optimizations, with only a modest 10% increase in computational time. These findings underscore the method’s effectiveness in balancing accuracy with resource efficiency.