Energy-Efficient Hyperparameter Optimization in Machine Learning Using Coati Optimization Algorithm (CMRLCCOA)

Abstract

Increasing demand for energy-efficient machine learning models requires optimization strategies that minimize computing costs and increase model performance. We use the Coati Optimization Algorithm to optimize SVM classifier hyperparameters. Lévy Flight explores, Sine Chaotic Mapping initializes populations, and Convex Lens Imaging Reverse Learning accurately searches. Iris dataset model accuracy and energy consumption improved with the method. CMRLCCOA balances computational economy and model accuracy, making it suited for energy-efficient machine learning.

CMRLCCOA-based SVM hyperparameter optimization improves exploration and exploitation through Sine Chaotic Mapping, Lévy Flight, and Convex Lens Imaging Reverse Learning. Machine learning models in resource-constrained situations must maximize accuracy and minimize energy use. We found that CMRLCCOA fits these standards. This study demonstrates that CMRLCCOA helps machine learning conserve energy while preserving model accuracy.

CMRLCCOA maximizes energy-efficient hyperparameter optimization by improving functional and non-functional SVM hyperparameters.