Абстрактный

Comparative Study of Different Classification Techniques for Post Operative Patient Dataset

Satya Ranjan Dash, Satchidananda Dehuri

Post Operative patient dataset is a real world problem obtained from the UCI KDD archive which is used for our classification problem. In this paper different classification techniques such as Bayesian Classification, classification by Decision Tree Induction of data mining and also classification techniques related to fuzzy concepts of soft computing is used for implementation of our dataset. The parameters used to comparison of different algorithms are RMSE, ROC Area, MAE, Kappa Statistics, time taken to build the model, Relative Absolute Error, Root Relative Squared Error, and the percentage value of classifying instances.