Explore high-quality datasets for your AI and machine learning projects.
The dataset contains transaction records of European credit‑card holders from September 2013. It comprises 284,807 transactions over two days, of which 492 are fraudulent. The dataset is highly imbalanced, with fraud (positive class) accounting for 0.172 % of all transactions. Only numerical input variables are provided, which are the result of a PCA transformation. The original features (Time and Amount) are not transformed. Time records the seconds elapsed between each transaction and the first transaction in the dataset. Amount is the transaction amount and can be used for cost‑sensitive learning. Class is the response variable, set to 1 for fraud and 0 otherwise.
