Post Your Answer
7 months ago in Machine Learning , Threat Intelligence By Kirti
Where to Get Good Data for Training an AI Intrusion Detection System
 I'm building a machine learning model to detect network intrusions. Where can I find reliable, labeled datasets to train and test my algorithm? I need something recent that reflects modern attack patterns.
All Answers (1 Answers In All)
By Shubham Answered 3 months ago
 The quality of your dataset makes or breaks an ML project. For intrusion detection, look at CICIDS2017 (current, diverse attacks) and UNSW-NB15. The classic KDD Cup 1999 and its improved version NSL-KDD are still used for benchmarking. For IoT-focused detection, check out TON_IoT. You can download most of these from places like Kaggle, the University of New Brunswick's website, or data.gov. Just remember to clean and understand the data before you start training.
Reply to Shubham
Related Questions