Comparative Analysis of Machine Learning and Deep Learning Approaches for Multiclass Identification and Classification of Alzheimer's Disease

Alzheimer's disease is a degenerative neurological condition which progressively impairs cognitive abilities and memory. Immediate prediction of Alzheimer's disease is crucial for timely intervention and treatment. This research paper presents a comprehensive approach for predicting Alzheimer's disease using Gray-Level Cooccurrence Matrix (GLCM) with machine learning (ML) models (K - nearest neighbour (KNN), Random Forest and Decision Tree), pretrained models(Xception, DenseNet-121, ResNet-50), and Convolutional Neural Network (CNN).The study utilizes a dataset obtained from Kaggle, which includes MRI images categorized into four different classes. The dataset is utilized to develop and assess different algorithms and techniques aimed at accurately predicting Alzheimer's disease. Texture features extracted from brain imaging data using GLCM are used as input for the machine learning models. Pretrained models are fine-tuned using transfer learning, and a CNN model is developed for direct feature learning. Performance of these models is compared, providing insights into their effectiveness for the prediction.