A New Harris Hawk Whale Optimization Algorithm for Enhancing Neural Networks

The learning process of artificial neural-networks is considered as one of the burdensome challenges to the researchers. The major dilemma of training the neural networks is the nonlinear nature and unknown controlling parameters like weights and biases. Slow convergence and trap in local optima are demerits of training neural network algorithms. To overcome these demerits, this work proposes a hybrid of Harris hawk optimization with a whale optimization algorithm to train the neural network. Harris hawk is a metaheuristic evolutionary algorithm and is used here to optimize the weights and bias of neural networks. The efficacy of the proposed algorithm is assessed by evaluating it on different kinds of cancer datasets and other datasets like fraud, banking note authentication. The experimental results demonstrate that the proposed algorithm performs better than its contemporary counterparts.