Abstract: - Due to rapid growth in the field of cashless transactions or digital transactions, credit cards are widely used in almost every work and hence there are more chances of fraudulent transactions. These fraudulent transactions can be identified by analyzing various behaviors of credit card customers from previous transaction history datasets. If any deviation is noticed in the behavior from the available patterns, there is the possibility of fraudulent transaction. Machine learning techniques are widely used to detect the frauds. In this paper, we have used KNN technique to detect the frauds .The performance of this techniques is evaluated based on the accuracy, sensitivity, precision and recall.
Keywords: - Machine learning, Credit Card, KNN technique, T-SNE technique, removing outliers, fraudulent