Breast Cancer Classification With Microarray Gene Expression Data Based on Improved Whale Optimization Algorithm

Breast cancer is one of the most common and dangerous types in women worldwide. Since it generally a genetic disease, microarray technology-based prediction technically significant among lot diagnosis methods. The gene expression data contains fewer samples with many redundant noisy genes. It leads to inaccurate diagnose low accuracy. To overcome these difficulties, this paper proposes an Improved Whale Optimization Algorithm (IWOA) for wrapper based feature selection data. proposed IWOA incorporates modified cross over mutation operations enhance exploration exploitation classical WOA. adapts multiobjective fitness function, which simultaneously balance between minimization error rate selection. experimental analysis demonstrated that, Gradient Boost Classifier (GBC) achieves high classification accuracy 97.7% minimum subset features also converges quickly breast dataset.