Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Optimization of Classification Results by Minimizing Class Imbalance in the Decision Tree Algorithm

Rektorat UDINUS 2022 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2022
Goals
Record
74800d08-9690-4bc1-880b-fae0a244a7a2
Published
2026-10-30

Universitas Dian Nuswantoro conducted a study on optimizing classification results by minimizing class imbalance in the Decision Tree algorithm. The activity involved Budiman, Fikri; Saputro, Irwan Agus; Purwanto, Purwanto; and Andono, Pulung Nurtantio as participants from UDINUS. The study addressed the imbalance in the number of training datasets for each class, a problem associated with classification boundary accuracy and unmaximal classification accuracy.

The activity examined the need to increase classification accuracy for minority data without causing a decrease in classification accuracy for majority data. To address the minority training dataset, the study applied a method to classification using the Decision Tree algorithm. The dataset used was Breast Cancer, with 286 instances and two classes. The no-recurrence-events class contained 201 instances, while the recurrence-events class contained 85 instances.

The ensemble classifier method applied to overcome class imbalance used a combination of AdaBoost, Bagging, and SMOTE algorithms. The study tested combinations of these algorithms and compared the resulting classifications. The combination of AdaBoost, Bagging, and SMOTE produced the highest accuracy value reported in the source, namely 84.3%. Through this activity, UDINUS conducted classification research using Breast Cancer data, minority and majority classes, the Decision Tree algorithm, and an ensemble classifier combination. The study's focus was the handling of class imbalance and the comparison of algorithm combinations.

Reference https://www.scopus.com/pages/publications/85128218825?origin=resultslist