Credit card transactions are one of the most business-critical problems because we deal with personal and critical information, and we also deal with the highly skewed nature of the dataset because fraud transactions are very smaller than natural transactions, so we deal with unbalanced data and target the small class, so it is a challengeable problem.
In this problem we want to compare the result of different models like traditional machine learning models, complex machine learning models and deep learning models
The data may be imbalanced and need some preprocessing so we will try hard to solve these different problems and build a good feature engineering and EDA. Imbalance data will be solved by two different techniques:
1- Random over sampling.
2- Random under sampling.