Comparative Analysis of Ant Colony Optimizer and Equilibrium Optimizer for Data Dimensionality Reduction in Image Processing and Pattern Recognition

Dimensionality reduction is a crucial step in handling high-dimensional data, as it helps extract essential information while mitigating the challenges posed by redundant or noisy features. In this study, we explore the effectiveness of two feature selection algorithms, Ant Colony Optimizer (ACO) and Equilibrium Optimizer (EO), for data dimensionality reduction in image processing and pattern recognition tasks. We apply these algorithms to a dataset containing the dimensions of electronic components’ footprints. The dataset comprises two subsets, feat length and feat_width, extracted from images of the length and width dimensions of the components, respectively. Through simulations, we compare the performance of ACO and EO in terms of accuracy and convergence speed. ACO demonstrates fast initial convergence but stagnates afterward, while EO exhibits slower initial convergence with continuous improvement. We evaluate their effectiveness by calculating the average accuracy and variance. The results show that ACO achieves a higher mean accuracy (93.03%) compared to EO (86.38%), but both algorithms perform well in finding good solutions.