Research on hybrid intrusion detection method based on the ADASYN and ID3 algorithms

Area of Science:

• Computer Science
• Cybersecurity
• Machine Learning

Background:

• Network security is crucial for preventing unauthorized access.
• Intrusion detection systems (IDS) are vital for early threat identification.
• Existing IDS methods require improvement in detection rates and accuracy.

Purpose of the Study:

• To propose a hybrid intrusion detection method combining ADASYN and ID3 decision tree.
• To enhance the effectiveness and accuracy of intrusion detection.
• To reduce the false alarm rate in network security.

Main Methods:

• Data preprocessing involved coding and normalization of the intrusion detection dataset.
• Adaptive Synthetic (ADASYN) algorithm was used for oversampling the training data.
• An ID3 decision tree model was developed for intrusion classification.

Main Results:

• The hybrid model demonstrated higher accuracy compared to other evaluated models.
• The proposed method achieved a lower false alarm rate.
• Performance metrics included accuracy, precision, recall, and false alarm rate.

Conclusions:

• The hybrid ADASYN and ID3 decision tree model is effective for intrusion detection tasks.
• This approach offers improved accuracy and reduced false alarms in network security.
• The method provides a more suitable solution for real-world intrusion detection challenges.