Credit Card Fraud Detection Model Based on Machine Learning

Credit card fraud can be defined as a situation in which an individual uses another person's credit card for an unauthorized transaction and neither the cardholder nor the card issuer is aware of the action. In this study, several credit card fraud detection models were constructed, and python language was used to implement the model construction, data dimensionality reduction and visualization. Through data analysis and visualization technology, the fraudulent transactions and non-fraudulent transactions of credit cards are distinguished. The data dimensionality reduction techniques used in this study include t-SNE dimensionality reduction, PCA dimensionality reduction and truncated SVD dimensionality reduction. The model construction techniques used in this study include random undersampling on the dataset, logistic regression, decision tree, KNN and support vector machine, providing an effective classification solution. The experimental results show that all the four models have better performance, and KNN has the highest training efficiency. Among the four models, decision tree classifier shows good performance while having good interpretability.