An Efficient Gradient Based Modified Cuckoo Search Optimization with Dynamic Levy Flight (MCSO-DLF) Algorithm for Breast Cancer Detection

• Published in: International Journal of Students' Research in Technology & Management Vol. 12; no. 2; pp. 1 - 12
• Main Authors: Chakraborty, Amit, Singh, Gulbir
• Format: Journal Article
• Language: English
• Published: 23.09.2024
• ISSN: 2321-2543; 2321-2543
• DOI: 10.18510/ijsrtm.2024.1221

Purpose of Study: This study evaluates the Gradient-Based Dynamic Levy Flight Cuckoo Search Optimization (MCSO-DLF) for breast cancer detection and compares it to six established optimization algorithms. The goal is to determine if MCSO-DLF offers higher accuracy and computational efficiency for optimizing diagnostic models. Methodology: The study compares MCSO-DLF with PSO, GA, ACO, SA, DE, and ABC using the Breast Cancer Wisconsin dataset. The algorithms are evaluated based on accuracy, convergence speed, and efficiency. MCSO-DLF uses dynamic Levy flight and gradient-based optimization to improve solution quality. Main Findings: MCSO-DLF achieved the highest accuracy (0.9912) and fastest computation time (65.94 seconds), outperforming all other algorithms. This demonstrates its effective balance between exploration and local solution refinement. Implications: MCSO-DLF significantly improves accuracy and speed, offering potential advancements in breast cancer detection systems, leading to better patient outcomes and more efficient healthcare. Novelty of Study: This study introduces MCSO-DLF as a novel hybrid optimization method that combines gradient-based optimization with dynamic Levy flight, outperforming traditional algorithms in complex medical diagnostics.