A classifier hub for imbalanced financial data

Abstract

We design and implement a classifier hub that can explore the detailed information on the imbalanced dataset and classify the dataset into two classes. Against the data imbalance, through setting imbalance ratio, it can adjust the proportion of majority and minority class. In this hub, we also implement Decision Tree, KNN and Random Forrest machine learning classifiers based on Python and Java. In the experiments, we use 30,000 loan records from an online P2P system as the dataset to demonstrate the functions of the classifier hub. The influences of different imbalanced ratio on classification performance have been compared through Decision Tree, KNN and Random Forrest algorithms.