XCOA-MLP: Extended Coyote Optimization Algorithm for training neural networks in medical data classification

Artificial neural networks (ANNs) are widely applied in medical data classification due to their ability to model complex and nonlinear patterns. However, their performance is highly contingent upon the effectiveness of their training methods. Traditional training methods such as backpropagation often suffer from slow convergence and local minima. Meta-heuristic algorithms provide global search capability but may face efficiency limitations. This study proposes XCOA-MLP, a multilayer perceptron trained using an improved coyote optimization algorithm (XCOA). XCOA extends the standard COA by introducing a leader pack (LP) technique that accelerates convergence and enhances exploitation through inter-pack knowledge sharing. The model was evaluated on five UCI medical datasets and benchmarked against eight meta-heuristic algorithms. Results show that XCOA-MLP achieves superior accuracy and robustness, recording 97.8% on Breast cancer, 78.2% on Diabetes, and 87.7% on Parkinsons datasets. These findings demonstrate XCOA-MLP’s effectiveness in improving neural network training for medical data classification.

Text 123912.pdf - Published Version Available under License Creative Commons Attribution Share Alike. Download (1MB) Official URL or Download Paper: https://inass.org/wp-content/uploads/2025/09/20251...

Actions (login required)

View Item