Analysis of Machine Learning Algorithms for Alzheimer’s Disease Classification and Prediction

EasyChair Preprint 9756

Abstract

Alzheimer's disease (AD) is a progressive neurodegenerative disorder in which dementia and cognitive decline gradually. AD is the major disease that is responsible for memory loss and cognitive impairment in older people. Till now there no known cure is available for AIAD. Early detection of AD and proper time treatment may prevent AD. Early detection of AD is possible using Machine Learning (ML) and MRI data. In this study, supervised ML classification along with longitudinal brain MRI data are used to predict and classify AD. The six states of art supervised classifiers named Random Forest (RF), XGBoost, Support Vector Machine (SVM), ExtraTree Classifier, Naive Bayes (NB) Classifier, and KNearest Neighbours (KNN) are used to classify and predict. The performance of the ML classifier was evaluated using performance metrics such as precision, recall, accuracy, f1-score and ROC/AUC curve. Among all six ML classifiers, the ExtraTree performed the best with the highest classification accuracy of 87%.

Keyphrases: Alzheimer’s disease (AD), Classification, Dementia, Performance, machine learning, prediction

@booklet{EasyChair:9756,
  author    = {Dhwani Modi and Pushpendra Singh Sisodia},
  title     = {Analysis of Machine Learning Algorithms for Alzheimer’s Disease Classification and Prediction},
  howpublished = {EasyChair Preprint 9756},
  year      = {EasyChair, 2023}}