Bagging Techniques to Reduce Errors in Parkinson^s Prediction Based on Decision Tree Irpan Kusyadi (*), Teti Desyani, Sri Mulyati, Endar Nirmala, Aries Saifudin, Yulianti
Universitas Pamulang
Jl. Raya Puspitek No. 46 Buaran, Serpong, Tangerang Selatan, Banten, Indonesia, 15417
*dosen00673[at]unpam.ac.id
Abstract
Parkinson^s disease is a neurological disease that is second only to Alzheimer^s disease and a complete treatment has not been found until now. It is necessary to slow the progression of this disease, as there is no cure for Parkinson^s. To slow the progression of Parkinson^s disease requires early diagnosis. However, the diagnosis of this disease cannot be made by laboratory tests. The traditional diagnosis is not suitable for early detection of Parkinson^s because it requires a lot of observation regarding daily activities, skills and other neurological parameters to assess the progression of Parkinson^s disease. Based on previous research, it has been found that Artificial Intelligence (AI) and Machine Learning have good classification potential and the classification system can improve the accuracy and reliability of diagnosis and also minimize errors and a more efficient system. The use of the Decision Tree has been used successfully for medical prediction and reliable decision-making techniques. The prediction results using the Decision Tree are still not perfect, so in this study, it is proposed to apply the bagging technique to reduce errors. Bagging is a simple but effective ensemble method that has been applied to many real-world applications. Bagging is an ensemble method that is widely applied to classification algorithms, to increase the accuracy of classifiers by combining single classifiers, and the results are better than random sampling. The experimental results show that the bagging technique can improve the performance of the Parkinson prediction model based on the Decision Tree.
