A Data-Driven Machine Learning Approach for Early Detection, Classification, and Progression Analysis of Alzheimer’s Disease

The Alzheimer's disease is a chronic, neurological and progressive disorder that affects the functions of memory, thought process and routine activities of a person. It is very essential to detect the Alzheimer's disease early for providing a better and effective patient treatment and care. In this paper we describe a machine learning based model to detect the Alzheimer's disease using tabular health data and Decision Tree classifier. The non-predictive features were eliminated and the target variable was encoded before scaling the numeric data to improve the performance of the classifier. We split the data in training and testing set with ratio 80:20. Then we train a Decision Tree Classifier model with max depth 5 to predict whether a patient has the disease or not. After predicting we evaluate the performance using accuracy score, confusion matrix, classification report and ROC curve analysis. From the experiment results we can see that the developed model gives the accuracy of 92.79% and AUC score of 0.93.