A Systematic Review of Machine Learning Models for Predicting Parkinson's Disease Progression

Parkinson Disease is a genetic illness which found primarily in Aged population i.e. after 60years. It is basically arising due to expiration of cell due to age in human brain, that causes degeneration of dopamine in Substantia Nigra Part (SNP). Symptoms of Parkinson’s disease are slow movements, face is of without expression, fingers of hand are vibrating, significant amount of changes in the voice of affected person, these symptoms of disease help to identify this disease, slow writing problem which is also helpful in identifying the disease. This disease cannot be healed but its early detection helps doctors to significantly slow down the growth of disease and also helps to reduce the symptoms of it. After reviewing several papers, it is identified that researchers had use models of areas like machine learning, Deep learning, Artificial intelligence etc., and apply this model on Parkinson’s disease affected person’s voice data or written data to diagnose the disease. Some uses optimization algorithm to increase the performance of model. . Some papers use unbalanced data, by applying some data balancing technique they balance the data.