Enhanced Employee Training Classification Model Using Optimized-HHO Probability Neural Network

Prominence has been given to Probabilistic Neural Networks (PNN) due to their proficiency in modeling intricate patterns and dependencies across diverse datasets. However, challenges such as slow convergence and susceptibility to local optima underscore the need for PNN optimization. This study focuses on elevating the efficiency and accuracy of PNN through the application of the Harris Hawk Optimization (HHO) algorithm. An Optimized-HHO approach is introduced to leverage global search capabilities while mitigating pitfalls of traditional HHO algorithms. This optimization process involves a fusion of the Golden Sinusoidal optimization algorithm during the exploration phase and a non-linear exponential decay strategy, enhancing global and local adaptability. The robust capabilities of Optimized-HHO are harnessed to refine the PNN's smoothing factor, achieving superior convergence precision. The paper presents an exploration of the construction and optimization methodology of the Probabilistic Neural Network using the Optimized-HHO algorithm (Optimized-HHO-PNN). Finally, application of Optimized-HHO-PNN to employee training needs classification is presented as a case study to highlight its practical utility and effectiveness. This research contributes to broader understanding of enhancing PNN performance through optimization techniques, with potential applications in various real-world scenarios.