The Parkinson’s Puzzle: Early Detection & Diagnosis

Parkinson's disease is a neurological disorder that affects movement and becomes progressively worse over time. Currently, there is no cure for the disease, but early detection and appropriate management can significantly improve the patient's quality of life. This project presents a web-based application designed to detect Parkinson's disease based on the analysis of audio recordings of patient's voices. The application utilizes a machine learning model that has been trained on a dataset consisting of audio recordings from both Parkinson's disease patients and healthy individuals. The web application extracts relevant features from the audio recordings, employing machine learning algorithms to predict the probability of Parkinson's disease. The machine learning models used in this research are Logistic Regression, Linear Discriminant Analysis, K-Neighbors Classifier, MLP Classifier, Gaussian NB, XGB Classifier, Random Forest Classifier, and Cat Boost Classifier and stacked three models (XG Boost, Random Forest, and Cat Boost) using Stacking CV Classifier out of which the model with the highest performance will be chosen to develop the web application. Keywords: Parkinson’s Disease, Web Application, Ensemble Learning, Stacked Model