An optimized artificial neural network approach based on sperm whale optimization algorithm for predicting fertility quality

Abstract

This research introduces a new approach for predicting reproductive health by using the Sperm Whale Optimization algorithm (SWA) with Artificial Neural Networks (ANN-SWA). SWA is an emerging method with a powerful potential in tackling optimization difficulties based on its adaptability in searching mechanisms. ANN-SWA consists of four phases. The first phase is conditioned by the fertility disease which is a complex and multifactorial condition with increasing incidence. The fertility data is trained (90 cases) and the approach is then used to test findings in the test set (10 cases). In the second phase, the multilayer perceptron (MLP) is used to determine the maximum fitness function by getting the global minimum and hence, it revokes the ANN trapped in local. In the third phase, it optimizes and controls the parameters (weights and biases) to ensure rapid convergence with accuracy. In the fourth phase, ANN-SWA is used to predict the fertility quality and determine the accuracy. The results are verified by comparing them with optimization and classification algorithms. The quantitative and qualitative outcomes show that the proposed approach is able to outperform the current algorithms on the fertility dataset in the convergence rate of classification. The results demonstrate that an artificial neural network based on SWA achieved more than 99.96% of the accuracy of the approach.