Using an improved wild horse optimizer - multilayer perceptron hybrid (IWHO-MLP) model in energy efficiency analysis

This study aims to develop strategies that can overcome local space entrapment and premature convergence problems in order to enhance the success of the Wild Horse Optimization (WHO) algorithm in optimization processes. Performance variation is observed through a Multilayer Perceptron (MLP) example. In this context, the Improved Wild Horse Optimizer (IWHO) algorithm has been developed using the Random Walk Strategy (RW) to provide solution diversity in local spaces. The challenging test suite CEC 2019 was selected to measure IWHO's performance. Its competitiveness was measured against alternative algorithms, demonstrating superior performance. This superiority has been visually demonstrated through convergence curves and box plots. The Wilcoxon signed-rank test was used to evaluate IWHO as a different and powerful algorithm. The IWHO algorithm addressed real-world problems by applying it to MLP training. Both WHO and IWHO algorithms were tested using MSE results and ROC curves. The Energy Efficiency Problem dataset from UCI (University of California Irvine) was used for MLP training. This dataset evaluates heating load (HL) or cooling load (CL) factors by considering the input characteristics of smart buildings. The objective is to ensure the most efficient evaluation of HL and CL factors through the use of HVAC (heating, ventilating, and air conditioning) technology in smart buildings. WHO and IWHO were selected to train the MLP architecture, and it was observed that the proposed IWHO algorithm produced better results.