Post-operative patients analysis using data mining techniques

Abstract

The post-operative patient analysis begins immediately after the surgery. The post-operative care depends on both the type of surgery and the previous health history. With the help of this analysis we can find if there exist any complications or not. In this paper we are analyzing the data of Post-Operative patients and we are identifying the state of the patient using J48 – a Data Mining Approach. After that, we are comparing various classification techniques in data mining and predicting their accuracy for post-operative patients data set. We are comparing Naïve Bayesian, SMO, LWL, J48 classifiers using performance measures like Receiver Operating Characteristic), Root mean squared Error, Kappa statistics and Mean Absolute Error using WEKA tool. Different Techniques in Data Mining could be used for the extraction of data from large data sources. Data Mining Approaches can be used in several fields like Medicine, Education, Fraud Detection, Marketing etc.