Parkinson's Disease Prediction using XGBoost and SVM

Abstract

Parkinson's disease or paralysis agitans or shaking palsy is a nervous condition that limits mobility. The condition may result in tremors, stiffness, or sluggishness of movement. Although there is no known treatment for this illness, the symptoms can be markedly reduced with medicines. To increase accuracy, this paper aims to forecast Parkinson's disease using machine learning methods like Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost). These algorithms perform better when used as classifiers, leading to higher accuracy. Hence using these algorithms, the machine learning model can easily predict Parkinson's disease. The accuracy of extreme gradient boosting was 92% and that of the support vector machine was 87%. XGBoost has a built-in capability to handle missing values and it provides various features. It is also effective with large datasets. Support Vector Machines outperform other classification algorithms in terms of performance and speed. It determines the best hyperplane for accurately classifying data points between several classes. The major goal of this study is to more accurately and effectively forecast Parkinson's disease so that it can be treated at the right time to minimize subsequent dangers. More Parkinson's disease cures and higher survival times result from early diagnosis. After detecting the disease, it can be cured through physical treatments and sometimes through surgery or in some cases through medications. It is observed that people diagnosed with Parkinson's disease are of an average age of 60, men are most likely to have this disease and it advances with the age of a person. The progress of Parkinson's disease is usually slow. The symptoms vary from person to person but are usually between a year or two.

Topics & Concepts