Comparative Analysis of Different Distributions Dataset by Using Data Mining Techniques on Credit Card Fraud Detection

Abstract

Banks suffer multimillion-dollars losses each year for several reasons, the most important of which is due to credit card fraud. The issue is how to cope with the challenges we face with this kind of fraud. Skewed "class imbalance" is a very important challenge that faces this kind of fraud. Therefore, in this study, we explore four data mining techniques, namely naïve Bayesian (NB),Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Random Forest (RF), on actual credit card transactions from European cardholders. This paper offers four major contributions. First, we used under-sampling to balance the dataset because of the high imbalance class, implying skewed distribution. Second, we applied NB, SVM, KNN, and RF to under-sampled class to classify the transactions into fraudulent and genuine followed by testing the performance measures using a confusion matrix and comparing them. Third, we adopted cross-validation (CV) with 10 folds to test the accuracy of the four models with a standard deviation followed by comparing the results for all our models. Next, we examined these models against the entire dataset (skewed) using the confusion matrix and AUC (Area Under the ROC Curve) ranking measure to conclude the final results to determine which would be the best model for us to use with a particular type of fraud. The results showing the best accuracy for the NB, SVM, KNN and RF classifiers are 97,80%; 97,46%; 98,16% and 98,23%, respectively. The comparative results have been done by using four-division datasets (75:25), (90:10), (66:34) and (80:20) displayed that the RF performs better than NB, SVM, and KNN, and the results when utilizing our proposed models on the entire dataset (skewed), achieved preferable outcomes to the under-sampled dataset.

Predictions and visualization for confirmed, recovered and deaths COVID-19 cases in Iraq

“…kNN can be defined as supervised learning approach that has been mainly utilized in the detection systems, also was extensively used in COVID-19 problems [21]. In such approach, new instance query has been classified on the basis of popular kNN distance measures, such as Minkowski distance, Euclidean distance, and Manhattan distance; we used the last one in this work.…”

Machine Learning for Credit Card Fraud Detection

“…Previous reviews indicate that the transaction aggregation strategies are primarily important in the functionality engineering process. The dataset [14] contains the transactions carried out by a card holder over a period of 2 days of january 2017. Where there are in total 569,614 transactions, including 984, or 0.172% of the transactions are fraudulent transactions.…”

“…In this study, we will use four well-known measures to assess the methodology, namely: precision, sensitivity, specificity and precision. These measures depend entirely on the four basic metrics (alarm rate), respectively "True Positive" (TP) of the number of fraudulent transactions which have been detected as a true alarm, "False Positive" (FP) the number of authentic transactions that were detected as a false alarm, "True Negative" (TN) the number of authentic transactions that were detected as a true alarm and "False Negative" (FN) are the number of fraudulent transactions missed [14], positive (P) means that the number of "fraudulent transactions" and negative (N) represents the number of "authentic transactions" the total of P and N means all transactions. Below are the equations for each measurement:…”

“…"The fraud detection in the credit cards is a binary classification which means that the transactions for a credit card are divided into two sets, either (0)-genuine transactions or (1)-fraudulent transactions. Using the confusion matrix for our study the calculation of alarm rates will be different [9], where the values for each alarm rate are taken from its position in the confusion matrix such as the TP, where the value was (1, 1), TN (0, 0), FP (0, 1) and FN (1, 0)" [14].…”

“…The confusion matrix summarizes the performance result of the algorithm as described on Table 3, it explains what the algorithm does correctly is what is not correct. The lines of the confusion matrix represent the predicted class, and that the lines represent the real class [14].…”