Implementation of xgboost for classification of parkinson’s disease

Abstract

Parkinson’s Disease (PD) is an advanced neurodegenerative illness. It is about 90% of PD sufferer shows speech disorders in the initial stages. Hence, in this research, speech features were applied to classify this illness. The most famous speech features used in PD research are jitter, shimmer, fundamental frequency parameters, harmonicity parameters, Recurrence Period Density Entropy (RPDE), Detrended Fluctuation Analysis (DFA), and Pitch Period Entropy (PPE).Those features were then called as baseline features used in this research. In this research, the XGBoost algorithm was used for the classification of PD. Initially, the whole baseline features were used in the XGBoost algorithm and obtained an accuracy score of the model 84.80%. For improving the model, feature selection was performed by plotting feature importance, which causes features of locShimmer (Fscore = 3) was excluded from the model. After feature selection was performed, the accuracy score of the model has increased to 85.60 %. We tried to improve the model using for second features selection, by excluding features with F-score values less than 20. However, after performed this feature selection, the accuracy of the model was decreased to 84.40 %. Thus, the model used is the model with an accuracy of 85.60%.