Summary of Detecting Anomalies in Blockchain Transactions Using Machine Learning Classifiers and Explainability Analysis, by Mohammad Hasan et al.
Detecting Anomalies in Blockchain Transactions using Machine Learning Classifiers and Explainability Analysis
by Mohammad Hasan, Mohammad Shahriar Rahman, Helge Janicke, Iqbal H. Sarker
First submitted to arxiv on: 7 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary As the popularity of Blockchain for digital payments grows, so does its susceptibility to malicious attacks. Anomaly detection within transactions is crucial for establishing trust in digital payments. However, this task is challenging due to the rarity of illicit transactions. Although studies have been conducted, a limitation persists: the lack of explanations for model predictions. This study integrates Explainable Artificial Intelligence (XAI) techniques and anomaly rules into tree-based ensemble classifiers for detecting anomalous Bitcoin transactions. The Shapley Additive exPlanation (SHAP) method measures feature contributions, while we present rules for interpreting transaction anomalies. Additionally, an under-sampling algorithm named XGBCLUS balances anomalous and non-anomalous data. This algorithm is compared to other techniques. Our results demonstrate that XGBCLUS enhances TPR and ROC-AUC scores, and our proposed ensemble classifiers outperform traditional single tree-based machine learning classifiers in terms of accuracy, TPR, and FPR scores. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Anomalies in Bitcoin transactions can hurt trust in digital payments. To stop this, we need to detect weird transactions better. But it’s hard because bad transactions are rare. Many studies have tried to solve this problem, but they didn’t explain why their models made certain predictions. This study changes that by combining two ideas: Explainable Artificial Intelligence (XAI) and tree-based machine learning algorithms. We used a special XAI method called SHAP to see which features in our data were most important. Then we created rules to help us understand what makes a Bitcoin transaction weird or not. We also made a new algorithm that balances our training data so it’s fair for both normal and abnormal transactions. Our results show that this approach works better than other ways of doing anomaly detection. |
Keywords
* Artificial intelligence * Anomaly detection * Auc * Machine learning
