An Extensive Survey of Various Deep Learning Approaches for Predicting Alzheimer’s Disease

Abstract

The mild cognitive impairments or Alzheimer’s disease (AD) prediction has captured the attention of various researchers due to the rapid growth of the disease and the necessity to treat AD in the earlier phase. However, due to the lack of high dimensionality neural data and the availability of lesser number of samples leads to the need for the better predictor model. The advancements in Artificial Intelligence (AI) paved the way for substantial growth of machine learning and deep learning approaches. The significance of deep learning approaches is comparatively higher than that of machine learning approaches in terms of prediction accuracy. Recently, deep learning is considered as the essential tool for predicting AD. Initially, the essential data have to be acquired from the proper source to perform the validation process. The features related with the disease progression needs to be analyzed and learned to reduce the dimensionality and computational overhead. Finally, the classification is performed with learning approaches to improve the prediction accuracy and works efficiently than the prevailing approaches. The anticipated model diminishes the need for human effort and provides easier way for intellectually diagnose the disease. Various experimentations are carried out to show the significance of diverse algorithm. The advantages and the disadvantages related to the prediction model is analyzed in an extensive manner.