The algorithm in machine learning plays an essential role in the identification of patterns in biomedical sciences. For complex medical problems, there are many classifications of medical images and forecasts of the model. The neurological disease which largely attacks the motor system are Parkinson Disease (PD). With the help of a Magnetic Resonance Imaging (MRI) or any kind of scan can be used which can detect and predict the disease. By using the level of the dopamine, the PD is detected. This paper presents an overview of various prediction and detection techniques used to identify Parkinson disease. The deep learning and machine learning algorithm are compared here. The survey contains ANN, CNN and neural network- based paper are compared. The different strategies and algorithms in disease prediction and detection are recognized and evaluated. The results and principal issues of each study paper are discussed and analysed in this paper.

Keywords : Parkinson Disease (PD),CNN ,ANN ,Dopamine ,Neural Network, Deep Learning, Machine Learning.