Credit Card Fraud Detection

Project information

Summary: Developed methodology to detect fraud transactions done using the credit card and determine which machine learning model works best on this type of dataset.

Tools: Python, NumPy, SciPy, Matplotlib, Scikit-Learn, and Pandas

Duties: It was an Individual project. I firstly collected and explored the dataset using various distribution charts and statistics, then balanced the dataset classes to avoid overfitting of the model’s performance, and then conducted modelling and prediction of data on different models.

Outcome: Detailed prediction and performance analysis was conducted at the end of the project and integrated the results using different charts developed using Matplotlib. I was successful in achieving 97% accuracy in the prediction of fraud and non-fraud transactions.

Result: Obtained High Distinction score for this project.