Accurate Credit Card Fraudulent Dataset using Logistics Regression Compared with Random Forest

Using the Logistic Regression technique, it is possible to forecast the accuracy % of credit card fraudulent transactions. For forecasting the accuracy percentage of credit card fraudulent transactions, Logistic Regression with sample size=100 and Random forest(RF) with sample size=100 were iterated with 95 percent confidence interval and pretest power of 80 percent at various periods. Because the sigmoid function used in logistic regression maps the value between 0 and 1, it aids in the prediction of accuracy % by improving the accuracy percentage prediction. When compared to the accuracy of Random forest, the accuracy of Logistic regression is much higher (98.2 percent) (92.4). A statistically significant difference existed between Logistic regression and RF, with (p=0.000) (p0.005) indicating a statistically significant difference. The Logistic Regression method aids in the prediction of credit card fraudulent transactions with more accuracy than other algorithms.