Voice Data Analysis for Early Detection of Parkinson's Diseaseusing Deep Learning Algorithms over Big Data

Parkinson's Disease (PD) is chronic and progressive movement disorder that affects the millions of people. It can grow continuously to halt the neural activities of PD affected people. The various researchers are designed prediction models to predict disease at early stage by analyzing various symptoms such as tremor, bradykinesia, postural instability and rigidity. These models are focused mainly on data analysis effectively to predict the disease in the initial stage to increase the patient life period using Machine Learning techniques. But, the present systems are not predicting the disease in time rigid by attaining multiple attributes on voice data set. The proposed system must be equipped with more characteristics for attaining multi-attribute Parkinson's symptoms analysis. In the proposed system, the Deep Speech Data Analysis (DSDA) is developed using Deep Learning algorithms. The DSDA based PD system can help to predict symptoms of PD effectively than the existing systems. The DSDA system includes the subsystems such as Deep Neural Network (DNN), Deep Recurrent Neural Network (DRNN), and Deep Convolutional Neural Network (DCNN). The DSDA is compared with existing works and showed better performance.