The Performance of the KMAR Conjugate Gradient Method in Training a Multi-layer Perceptron Neural Network for COVID-19 Data
摘要
In this chapter, a new conjugate gradient MethodConjugate gradient, called KMAR, is employed for solving unconstrained optimization problems, specifically to train a multi-layer perceptron (MLP) neural network. The aim is to develop a resilient model capable of forecasting diverse Covid-19 cases Covid-19 in Malaysia employing a time series methodology. The numerical results affirm the remarkable efficacy of the proposed method, outperforming traditional approaches.