MACHINE-LEARNING AND DEEP-LEARNING BASED PARKINSON DISEASE DETECTION SYSTEM

MACHINE-LEARNING AND DEEP-LEARNING BASED PARKINSON DISEASE DETECTION SYSTEM

Skanda G N*1, Dr. Sankhya N Nayak*2, Sakshi M*3, Pavana R*4, Shreyas R Adiga*5

*1,3,4,5Research Scholar, Dept. of CS&E, JNNCE, Shimoga, India

*2Associate Prof., Dept. of CS&E, JNNCE, Shimoga, India

ABSTRACT

As Parkinson Disease patients show unique and characteristic vocal features and change in writing patterns, their voice recordings and drawings would prove to be an invaluable procedure for diagnosis as it is also non-invasive. The work on “Machine-learning and Deep learning-based Parkinson disease detection system” aims to provide a web application and a system for diagnostic purpose in hospitals or any testing centres. Machine Learning and Deep learning Algorithms are used to identify Parkinson’s disease. The status provided to the users will be displayed based on accuracy of multiple machine learning algorithms for voice dataset and Convolutional neural network for spiral dataset. Therefore, this work aims to provide a convenient and non-invasive way of predicting Parkinson's disease.

Keywords: Parkinson Disease, Convolutional neural network