Ensembling and Modeling Approaches for Enhancing Alzheimer's Disease Scoring and Severity Assessment

With an increasing elderly population and the expected surge in Alzheimer's disease (AD) cases, there is a pressing need for efficient AD detection tools. This study harnessed natural language processing and machine learning for predictive modeling of the Mini-Mental State Examination (MMSE). Our LightGBM model achieved state-of-the-art results, surpassing existing benchmarks by 37 % in test RMSE. We also introduced an ensemble strategy, which further enhanced performance by 60 %. Our novel AD severity classification approach also showed lightgbm and SVM as top performers. These advancements underscore significant strides towards practical AD detection applications.