Comparative Analysis of Data Mining Models for Classification for Small Data Set

The aim of this research paper is to identify best model to classify a data set with target variable as mode of delivery of pregnant women and a set of predictors (Mother’s weight, height, age, baby’s weight, and baby’s gender). The analysis is carried out using different algorithms of classification taking one at time which are part of Weka and R-Programming library. In addition, ensemble technique having different base classifiers is also applied to the same data set. The base classifiers used for ensemble techniques are Simple Cart, C4.5, Multi-Layer Perceptron (MLP), Radial Bias Function (RBF), and Reduced Error Pruning Tree (Reptree). The results with respect to performance of the new models are compared with the performance of logistic regression, discriminant function, and standalone decision tree algorithms based on values of accuracy, precision and recall [1]. It is concluded that data small or big should be subjected to many algorithms and their combinations using either hybrid or ensemble and other existing approaches, to get a more reliable output.