Akademik Veri Yönetim Sistemi
18th International Conference on Information Security and Cryptology (ISCTürkiye), Ankara, Türkiye, 22 - 23 Ekim 2025, ss.1-5, (Tam Metin Bildiri)
This study aims to classify the “specific result” output, a rarely explored aspect in the literature, using a widely adopted dataset in SCADA system security. While most previous studies have concentrated on binary or categorized result outputs, this research uniquely focuses on specific result classification. To mitigate class imbalance and enhance model performance, the Synthetic Minority Oversampling Technique (SMOTE) was applied. Subsequently, multiple machine learning and deep learning algorithms, including Random Forest, Logistic Regression, Decision Tree, Naive Bayes, Support Vector Machine, Convolutional Neural Network, Long Short-Term Memory, Extended Long Short-Term Memory, Multilayer Perceptron, Deep Neural Network, and Gated Recurrent Unit, were implemented. Experimental results demonstrated that the Random Forest model achieved an accuracy of 99%, while the MLP model reached 95%. These results outperform previous studies on specific result classification and confirm the effectiveness of the proposed approach for attack detection in Industrial Control Systems.