Performance Analysis of SMOTE-Enhanced Machine Learning Approaches on IoT-Based TON IoT Dataset Nesnelerin Interneti Temelli TON IoT Veri K mesinde SMOTE ile G lendirilmis Makine grenimi Yaklasimlarinin Performans Analizi

Kalifati E., Şolpan Ş., Eren K. K., Küçük K.

33rd IEEE Conference on Signal Processing and Communications Applications, SIU 2025, İstanbul, Türkiye, 25 - 28 Haziran 2025, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu66497.2025.11111750
  • Basıldığı Şehir: İstanbul
  • Basıldığı Ülke: Türkiye
  • Anahtar Kelimeler: Data Imbalance, Internet of Things (IoT), KNN, Machine Learning, SMOTE, SVM
  • Kocaeli Üniversitesi Adresli: Evet

Özet

Internet of Things (IoT) devices, despite their wide range of applications, often produce datasets with imbalanced class distributions. This imbalance negatively impacts the accuracy of intrusion detection systems. In this study, the Synthetic Minority Over-sampling Technique (SMOTE) was employed to address the class imbalance in data collected from IoT devices. By artificially generating minority class samples, the performance of the K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) algorithms was evaluated. Experiments were conducted on the ToN_IoT dataset, which includes data from seven different IoT devices, and performance was assessed using metrics such as accuracy, precision, and F1-score. The results demonstrate significant improvements in accuracy, particularly for Modbus and motion light devices. KNN showed greater improvement after applying SMOTE, achieving an accuracy of 99.9% for the weather device. This study highlights the effectiveness of SMOTE in mitigating data imbalance in IoT applications.