Comparative Study on Parkinson Disease Dignosis Treatment Classification Using Machine Learning Classifier (PDMLC)

Parkinson’s infection is a cerebrum issue that prompts shaking, solidness, and trouble with strolling, parity, and coordination. Parkinson’s side effects for the most part start step by step and deteriorate after some time. As the infection advances, individuals may experience issues strolling and talking. Parkinson disease is a neurological issue, which is one of the most excruciating, risky, and nondurable infections, which happens at more established ages. The Static Spiral Test (SST), Dynamic Spiral Test (DST) and Circular Motion Test(CMT) on certain point records were utilized in the analytical modelling application which was created for the conclusion of this malady. The dataset is collected from the online authorized source to classify the test basis on the symptoms. These datasets were separated into 70–30% of splitting into the training and testing information inside the original and control structures of ten-fold cross-validation approval method. Training set is utilized for train the ML-based analytical classification models such as logistic regression, neighbors classifier, decision tree, and support vector machine. Model is also trained by performing hyper-parameter tuning. After training phase, the model is assessed with test and the comparative results are discussed. Additionally, new information investigation was completed. As indicated by the outcomes acquired, SVM with RBF kernel is more effective than other classifier; furthermore, logistic regression calculations are in investigation of new information. This study contributes in the same direction by analyzing the behavior according treatment classification.