Data-Driven Prescriptive Modelling using Optimized Deep Learning

Abstract

This paper proposes an optimized deep learning algorithm for predictive analysis of health care data. The proposed algorithm is a hybrid model of a convolutional neural network and a moth-flame optimization algorithm. The moth-flame optimization algorithm reduces the multi-variant features of the healthcare datasets. the optimized features of the moth flame optimization process in a convolutional neural network. The hybrid model increases the performance of predictive parameters. The proposed algorithm was simulated in MATLAB software and tested on reputed datasets of heart diseases. The performance of the proposed algorithm compares to existing algorithms in deep learning such as CNN, RNN, and GAN. The performance of the proposed algorithms suggests that proposed are very efficient compared to existing algorithms for deep learning