The Ethereum network, is subject to a number of assaults from diffrent kinds. All of these frauds make it mandatory to have intrusion detection system (IDS) on blockchain networks. However, such a system requires mining a (big enough) dataset, labeling it, and making a good classifier. We will try in the following project to cover all these steps. We will try to answer the following questions:
- Are scammers databases enough to create anomaly detection models ?
- Which classifier works the best for this task?
- Which features are best determining the fraudulent nature of a transaction, in this context?
Please refer to the pdf report to see details about the workflow and the results.
- Hatem Mnaouer
- Salma Ezzina
- Amal Zili