Nowadays, the banking sector is a very necessary sector in our present day generation where almost each and every human has to deal with the bank either physically or online. In dealing with the banks, the clients and the banks face the possibilities of been trapped by fraudsters. Examples of fraud include insurance fraud, credit card fraud, accounting fraud, etc. The most frequent payment mode is credit card for both on-line and offline in today’s world, it presents cashless purchasing at each store in all countries. It will be the most convenient way to do on-line shopping, paying bills etc. Hence Detection of fraudulent activity is necessary in credit card transactions. Credit card fraud can be efficiently detected by data mining techniques. Data mining techniques such as classification, clustering and association rules analyze the credit card transaction data and predict the patterns which lead to fraud. Because of increased credit card fraud, the main aim of this research is to develop a model that should efficiently analyze and predict the fraudulent data set. In this research, an efficient credit card fraud detection model P-SVM classification is proposed and compared with decision tree and Naïve Bayes algorithms. The proposed model uses credit card holder’s expenditure behavior and location analysis for analyzing and predicting fraud in the credit card transaction data. Various evaluation measures are used for analyzing the performance of Naïve Bayes, Decision Tree and P-SVM classification technique and proved that P-SVM yields higher accuracy when compared with Naïve Bayes and Decision Tree.