Analyzing Dementia Prediction Models with Deep Neural Networks on OASIS Data

Abstract

This study presents a comprehensive investigation into the early prediction of dementia using longitudinal data, focusing on the evaluation of various machine learning classifiers. In the methodology, we meticulously prepare and preprocess the data, including data visualization, imputation, transformation, and feature selection. Subsequently, we apply an array of hyper-parametric classifiers, including logistic regression, linear discriminant analysis, k-nearest neighbors, decision trees, naive Bayes, support vector machines, ensemble methods like random forest and XGBoost, among others. The performance of these classifiers is rigorously assessed, with random forest and XGBoost emerging as the top performers, achieving accuracy rates.