Machine Learning Based Prediction Alzheimer’s Disease Using RFC-LSTM

Abstract

Based on the idea that a country's progress starts with improving the performance of its community-serving institutions, such as the Ministry of Health, and in light of technological advancements and the growing need to prevent various diseases and identify diseases that affect the elderly, like Alzheimer's disease, it has been discovered that the world has recently resorted to smart data analysis techniques and spatial deep neural computing in the healthcare industry to predict high-quality results quickly. A model for Alzheimer's patient prediction utilizing multivariate analysis and deep neural computing optimal technology (LSTM-RFC) is presented in the study. There are five fundamental stages in this model: The first step involves gathering data and preparing it for the decision-making stage. This includes a number of stages, such as processing missing numbers and modifying the goal. Creating methods to create an ideal structure for one of the deep brain computing networks—long short-term memory, or LSTM—is the second step. This tool was chosen following a thorough analysis that focused on identifying the main programming processes, important parameters, and the benefits and drawbacks of each method in order to achieve optimization techniques such as PSO, BOA, WOA, COA, and FA. The optimal architecture for LSTM, a deep neural computing technology, is determined through optimization approaches. which, out of several technologies (including recurrent neural networks (RNN), gated recurrent units (GRU), long short-term memory (LSTM), bidirectional long short-term memory (BiLSTM), AlexNet, and GoogleNet), was chosen for development following Camper. Because WOA is an algorithm with various benefits and features, this camp operates based on the programming processes and important parameters impacting each algorithm. The suggested approach seems to be a useful intelligent data analysis model that can cut down on the time and processing required to handle large, real-world data.