Accuracy based Performance Analysis of Classification Algorithms using Parkinson Disease Dataset

Nowadays, a variety of Parkinson's Disease (PD) has become common among humans which unfortunately lead the cause of death for many peoples in the globe. The chances of survival and the probability of receiving a cost-effective treatment can increase hugely with early finding and diagnosis of the ailment. Speech impairment and difficulty in performing the day-to-day activities are the most common symptoms for the progress of the disease. Naïve Bayes, k-Nearest Neighbor (k-NN), Support Vector Machine (SVM), and Random Forest Different are the Machine Learning (ML) classifiers that were used for classification of PD data set. When it comes to investigating an unclear data from a particular clinical informational collection, Clinical data mining is very efficient. The presentation of the technique has been surveyed with a trustworthy dataset from UCI repository. This paper expects current techniques of information disclosure in data sets that are being used today for the classification of Parkinson Disease. A comparative study on various classification models is done to the chosen dataset by applying the feature selection analysis and the accuracy analysis to come up with the best classification rule. Likewise, the point of the review is to channel the information in a manner that the healthy people and people with Parkinson are grouped accurately. The Random Forest algorithm performs well for classifying the information in view of the precision through its performance.