Early Parkinson Disease Detection With Feature Extraction Using Machine Learning

Abstract

Parkinson's disease (PD) is a neurodegenerative disorder impacting millions worldwide. This study aims to leverage machine learning algorithms to improve the diagnosis and understanding of PD. Building upon existing research that utilizes classifiers, feature extraction, data partitioning, and feature selection, we explore the potential of various feature selection algorithms in maximizing classification accuracy on publicly available PD datasets. The investigation will compare and contrast the performance of feature selection technique, ultimately identifying the method that yields the highest accuracy in PD classification. This research contributes to the growing body of knowledge surrounding the application of machine learning in PD diagnosis and paves the way for further exploration of specific feature sets and classification models to enhance clinical practice and patient outcomes. Here we will work on different selecting algorithms. Such as:PCA. For the result we will use several popular ML techniques. Including: K-NN, Random Forest, Decision Tree Algorithm,XG Boost,ANN.After our study of parkinson disease classification with the feature selectionwith correlation method we got accuracy of 89% on the Random Forest and the result after applying the PCA classifier the accuracy increased to 92%.