Predictive Analytics Algorithms for Clinical Decision Making in Healthcare

Abstract

Healthcare is major issue and challenge now-a-days for human being in a daily life. Parkinson Disease or P.D is a one of the disorders that affected in the mid of nervous system. Parkinson Disease affected person cannot be act as normal human being. Among the innumerable disease listed so far, the Parkinson disease occupies an alarming position due to its life threaten concern. Early prediction of Parkinson disease from large volume of electronic health records leads to protect various health issues. There are various challenges and issues such as scalability, accuracy, risk factor, time complexity and sparsity in early prediction of Parkinson disease. There are various conventional algorithms have been proposed to solve these issues and challenges and still needs improvement. The present study, systematic predictive analytics using various classification algorithms such as Support Vector Machine (SVM), Random Forest, AdaBoost, Multi-Layer Perceptron (MLP), Naive Bayes, Decision table, J48, Logistic Regression is presented and evaluated using benchmarking Parkinson disease data set which are collected from UCI machine learning repository. The extraction of hidden data present in the dataset is obtained using WEKA environment. The results from the prediction models gives better clinical decision-making support to the doctors in predicting disease earlier and risk level.