10.22376/ijpbs.2019.10.1.p1-12 Volume 6 Issue 1 2015 (January - March) Mining the data sets of different sizes or different regions many times will not yield expected maximum accuracy. Hence the data size becomes an important parameter for mining exercises. In this paper, we consider data from two different geographical regions and calculate separate performance measures. Also, we get the same for integrated data set obtained by the union of the original sets as inverse results establishing the hypothesis for integrated data set. R.KARTHIKEYAN, A.KUMARAVEL AND A.CHANDRA SEKAR Heart disease data sets decision trees, decision rules, Meta classifiers, Bayes classifiers. 177-187