Prediction of default credit card users using data mining techniques

Development of financial sector has lead to an increase in financial risk. In order to prevent such risks, this study proposes a model for prediction of default cards with the help of data mining techniques. Balancing algorithms such as SMOTE and ADASYN algorithms are used to balance the imbalanced data because balanced data can be useful in increasing the efficiency of the model. Later both the balancing techniques are compared to see which one performs better. This balanced data is then taken as an input to an machine learning algorithm such as SVM to predict default credit cards. Accuracy of this model is found out by comparing it with other data models.