SIRES 2019 Conference

Machine Learning Approach to Predict and Evaluate Banking-s Business Performance and Bankkruptcy
1st Bambang Siswoyo Information System (of Affiliation) Maksoem University (of Affiliation) Jl. Raya Cipacing No. 22, Indonesia; 2nd Nanna Suryana FTMK (of Affiliation) Universiti Teknikal Malaysia Melaka (of Affiliation) Melaca, Malaysia;3nd Zuraida FTMK (of Affiliation) Universiti Teknikal Malaysia Melaka (of Affiliation) Melaca, Malaysia



Abstract—The prediction of a companys bankruptcy, especially the prediction of bankruptcy of the banking industry in Indonesia, is very important. With this bankruptcy information, Bank Indonesia (BI) as the central bank, is able to make policies and develop a sound banking system. This research has been developed to build a model that is able to predict and evaluate the bankruptcy of the banking industry. The dataset used is the published financial ratio, for the algorithm used is the twoclass boosted decision tree. Machine learning learning with a dataset of input-output models can be implemented between financial ratio variables against bankruptcy. Overall, the model used is able to train input- output relationship data and system behavior properly. Overall, it is able to train and model input-output relationship behavior well, with AUC result of 95% and 95% accuracy, Precision 95% and 100% recall.

Keywords: Keywords—Machine learning, bankruptcy, twoclass decision tree

Topic: Other Related Topics


Web Format | Corresponding Author (Bambang Siswoyo)