Cryptocurrency blockchain data encounter a class-imbalance problem due to only a few known labels of illicit or fraudulent activities in the blockchain network. For this purpose, we seek to compare various resampling methods applied to two highly imbalanced datasets derived from the blockchain...