Comparison of Novel Optimized Random Forest Technique and Support Vector Machine for Fraudulent activities in credit card Detection with Improved Precision

The objective in the research work is to determine fraudulent activities in credit card using Novel Optimized Random Forest Technique (NORFT) algorithm with comparison of Support Vector Machine (SVM) algorithm by improving the precision. The study contains two groups, i.e Novel Optimized Random Forest Technique Algorithm is developed in the first group and Support Vector Machine is developed in the second group. To categorize data, a sample size of 181 per group was used with a g-power of 80%. The data was collected from various recent studies on the web, using a threshold of 0.05%, a confidence interval of 95%, and mean and standard deviation measurements. A new algorithm called NORFT was compared to the SVM, and NORFT was found to have significantly higher precision at 92.52%, compared to SVM Algorithm's precision of 62.82%. The significance value (p>0.05), which was determined to be p=0.352, indicated that there was no significant difference between the two algorithms. However, NORFT was found to be more accurate in predicting fraudulent activities in credit card and was able to improve precision levels compared to SVM.