IBGJO: Improved Binary Golden Jackal Optimization with Chaotic Tent Map and Cosine Similarity for Feature Selection

Feature selection is a crucial process in machine learning and data mining that identifies the most pertinent valuable features dataset. It enhances efficacy precision of predictive models by efficiently reducing number features. This reduction improves classification accuracy, lessens computational burden, overall performance. study proposes improved binary golden jackal optimization (IBGJO) algorithm, an extension conventional (GJO) algorithm. IBGJO serves as search strategy for wrapper-based feature selection. comprises three key factors: population initialization with chaotic tent map (CTM) mechanism exploitation abilities guarantees diversity, adaptive position update using cosine similarity to prevent premature convergence, well-suited problems. We evaluated on 28 classical datasets from UC Irvine Machine Learning Repository. The results show CTM based proposed can significantly improve Rate convergence GJO accuracy also better than other algorithms. Additionally, we evaluate effectiveness performance enhanced factors. Our empirical help algorithm converge faster.