Missing Data Imputation in Large Data set Using Chernaïve Classifier

Abstract

ABSTRACT
Data mining is a knowledge domain for information industries and communal sites by abstracting and refining the information from massive architecture. The Objectives of the study highlights the transformation of the limitations of Multiple Imputation in Large Data set through Adaptive boosting Algorithm, Naïve Bayesian (NB), J48 algorithm, CHAID and to construct Chernaïve Classifier to enable Missing Data Prediction on defining the large data set. This research aims to prove that Chernaïve Classifier overcomes the limitations of Decision Tree (DT), Adaptive Boosting (ADAB), Naive Bayesian and J48. A Mathematical model is constructed implementing the features of Chernaïve Classifier. This model overcomes the issue of independence classifier and boosting techniques, to implement the prediction of missing data in the historical data items. To implement every stages of the research work standard expertise tools like MATLAB, and SPSS for evaluation were used.

Keywords: Chernoff Bounds, Naive Bayes classifier, Decision Tree, Adaptive Boosting, J48, Missing Data Imputation, Classifier, Chernaïve Classifier