Improving imbalanced datasets in machine learning with synthetic data

Use synthetic data and to improve model accuracy for fraud, cyber security, or any classification with an extremely limited minority class.