WoM-based deep BiLSTM: smart disease prediction model using WoM-based deep BiLSTM classifier

Diagnosis of cardiovascular disease has been significant due to the increased number of people affected by cardiovascular diseases. Though various methods were developed in classifying the diseases and ensuring the privacy for secure data transfer, most of the existing methods suffer from accurate decision making. Hence, this research tends to introduce a well-suited disease prediction model with the help of an improved deep Bidirectional Long Short Term Memory (Deep BiLSTM). The hyper-parameters related to the optimized deep BiLSTM classifier are tuned by the proposed optimization named Whale-on-Marine optimization (WoM) algorithm. The research-based on the developed deep BiLSTM classifier for heart disease utilizes the Elliptic Curve Cryptography (ECC) dependent Diffi-Huffman algorithm to ensure secure data transmission in the network. The effective decision making and secure data transfer are performed using the WoM-based deep BiLSTM classifier, where the WoM optimization reflects the characteristics of marine and Whale to determine the space between the prey and the predators, which improves the exploitation integration and exploitation tendencies. The performance metrics reveal that the proposed optimization based on deep BiLSTM effectively predicts heart diseases. The optimized deep BiLSTM classifier achieves a sensitivity of 97.93%, specificity of 97.52%, F-Measure of 97.658% and accuracy of 97.53% for the training percentage in Statlog, Cleveland, and Hungary database.