Dealing with the class imbalance issue in the data is a big problem when creating fraud detection systems since valid
transactions exceed fraudulent transactions by a large degree. Fraudulent transactions generally make up less than 1% of all
transactions.