Improved Reptile Search Optimization Algorithm: Application on Regression and Classification Problems

The reptile search algorithm is a newly developed optimization technique that can efficiently solve various problems. However, while solving high-dimensional nonconvex problems, the retains some drawbacks, such as slow convergence speed, high computational complexity, and local minima trapping. Therefore, an improved (IRSA) based on sine cosine Levy flight proposed in this work. modified with enhanced global capabilities avoids trapping by conducting full-scale of solution space, operator jump size control factor increases exploitation agents. was applied to set 23 well-known test functions. Additionally, statistical analysis performed considering 30 runs for performance measures like best, worse, average values, standard deviation. results showed gives fast low time efficient search. For further verification, were compared RSA state-of-the-art metaheuristic techniques. In second phase paper, we used IRSA train hyperparameters weight biases multi-layer perceptron neural network smoothing parameter (σ) radial basis function network. To validate effectiveness training, trained classifier tested challenging, real-world classification Furthermore, application, IRSA-trained RBFNN regression model day-ahead wind solar power forecasting. Experimental clearly demonstrated superior prediction hybrid model. Qualitative, quantitative, comparative, statistical, complexity revealed exploration, efficiency, accuracy, technique.