Early Alzheimer’s Disease Prediction Using Machine Learning
Aryan Karve
Dept. of Electronics and Telecommunication Engineering
PES’S Modern College of Engineering
Pune, India
aryan7karve@gmail.com
Aarya Dhaygude
Dept. of Electronics and Computer Engineering
PES’S Modern College of Engineering
Pune, India
aaryamdhaygude18@gmail.com
Vedant Kugaonkar
Dept. of Artificial Intelligence and Data Science Engineering
PES’S Modern College of Engineering
Pune, India
vedant.kugaonkar@gmail.com
Prof. Prajakta pardeshi
Department of Electrical and Electronics Engineering
Dr. Vishwanath Karad MIT World Peace University
prajakta.pardeshi@mitwpu.edu.in
Abstract—Alzheimer's Disease (AD) is a neurodegenerative condition that progressively erodes cognitive functions, especially in the older population. Early detection is incredibly crucial to treat symptoms and prevent disease progression. This paper presents a machine learning-based classification system for early prediction of AD using Decision Tree (DT), Random Forest (RF), and Gradient Boosting (GB) algorithms. The dataset is preprocessed for feature selection, normalization, and class balancing from Alzheimer's Disease Neuroimaging Initiative (ADNI) and OASIS. The input features used for model training and testing include structural MRI features, cognitive test results, and biomarker values. Experimental results indicate ensemble models like Random Forest and Gradient Boosting perform better than conventional methods in precision, recall, F1-score, and classification accuracy rate of up to 89% in classification accuracy. Explainability AI (XAI) techniques like SHAP and LIME enhance explainability and enable clinicians to understand significant factors influencing predictions. The method demonstrates that machine learning with explainability supports timely diagnosis and sound clinical decision-making. Additional research employs longitudinal and multimodal data to provide more robust models and apply them in everyday environments.
Keywords—Alzheimer’s Disease (AD), Machine Learning, Deep Learning, Early Diagnosis, Random Forest, Support Vector Machine, XGBoost, Convolutional Neural Network, Explainable AI, Feature Selection, Neuroimaging.
