Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Jurnal Teknik Informatika (JUTIF)
Bangun Watono
Magister of Informatics Engineering, Universitas AMIKOM Yogyakarta, Indonesia
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
STACKING ENSEMBLE LEARNING AND INSTANCE HARDNESS THRESHOLD FOR BANK TERM DEPOSIT ACCEPTANCE CLASSIFICATION ON IMBALANCED DATASET Bangun Watono; Ema Utami; Dhani Ariatmanto
Jurnal Teknik Informatika (Jutif) Vol. 5 No. 4 (2024): JUTIF Volume 5, Number 4, August 2024 - SENIKO
Publisher : Informatika, Universitas Jenderal Soedirman

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52436/1.jutif.2024.5.4.2022

Abstract

Bank term deposits are a popular banking product with relatively high interest rates. Predicting potential customers is crucial for banks to maximize revenue from this product. Therefore, bank term deposits acceptance classification is an important challenge in the banking industry to optimize marketing strategies. Previous studies have been conducted using machine learning classification techniques with the imbalanced Bank Marketing Dataset from the UCI Repository. However, the accuracy results obtained still need to be improved. Using the same dataset, this study proposes an Instance Hardness Threshold (IHT) undersampling technique to handle imbalanced datasets and Stacking Ensemble Learning (SEL) for classification. In this SEL, Decision Tree, Random Forest, and XGBoost are selected as base classifiers and Logistic Regression as meta classifier. The model trained on SEL with the dataset undersampled using IHT shows a high accuracy rate of 98.80% and an AUC-ROC of 0.9821. This performance is significantly better than the model trained with the dataset without undersampling, which achieved an accuracy of 90.30% and an AUC-ROC of 0.6898. The findings of this research demonstrate that implementing of the suggested IHT undersampling technique combined with SEL has been evaluated to effectively enhance the performance of term deposit classification on the dataset.
Co-Authors Dhani Ariatmanto Ema Utami