{
  "fetched_at": "2026-10-20T00:13:27Z",
  "source_url": "https://www.kaggle.com/competitions/ddb-inhouse-day-challenge/data",
  "pages": [
    {
      "name": "data-description",
      "content": "Overview of the provided data files\n\n## Files\n\n*   **train.csv** - the training set\n*   **test.csv** - the test set\n*   **sample_submission.csv** - a sample submission file in the correct format\n\n## Columns\n\n- `txid` — the transaction identification number\n- `creationdate` — the transaction date\n- `issuercountrycode` — the country where the shopper’s card was issued\n- `txvariantcode` — the card type\n- `bin` — the first four to six digits of the card number are called as the Bank Identification Number (BIN), they can be used to quickly determine the card network and the financial institute that issued the card\n- `amount` — the transaction amount in the original currency\n- `currencycode` — the currency the transaction was processed in\n- `shoppercountrycode` — the shopper country represents the location of the shopper during the transaction and is recorded based on both the issuing country of their payment method, as well as the IP address at time of transaction\n- `cardverificationcodesupplied` — cvc code supplied\n- `cvcresponsecode` — response code from the issuer concerning the cvc code\n-  `card_id` — a hashed version of the card number\n- `is_fraud` — The target column indicating whether a payment is fraudulent or not",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "For this challenge, we expect you to come up with two results:\n\nProbability for the payments in the test dataset (fraudulent or non-fraudulent) [0, 1]\nA presentation of maximum 5 minutes in which you describe the problem, how you tackled the fraud, and what your recommendations are to the Adyen board\n\n## Key points\n\n- Merchants really dislike it when non-fraudulent payments are blocked because the model classified them as fraudulent payments\n- Launch fast and iterate: don’t spend too long exploring the data, but try to create an end to end pipeline as quickly as possible. Once you have this in place you can start trying to improve your results. Make sure you keep track of any incremental improvements.\n- Make sure to share evaluation metrics (e.g., recall, accuracy, F1) in your presentation\n- Be creative and come with an innovative, out-of-the-box solution and presentation, we like to be different from our competitors\n- Convincing us why your solution is the best is just as important as having the best solution\n\n## Definitions\n\n- `txid` — the transaction identification number\n- `creationdate` — the transaction date\n- `issuercountrycode` — the country where the shopper’s card was issued\n- `txvariantcode` — the card type\n- `bin` — the first four to six digits of the card number are called as the Bank Identification Number (BIN), they can be used to quickly determine the card network and the financial institute that issued the card\n- `amount` — the transaction amount in the original currency\n- `currencycode` — the currency the transaction was processed in\n- `shoppercountrycode` — the shopper country represents the location of the shopper during the transaction and is recorded based on both the issuing country of their payment method, as well as the IP address at time of transaction\n- `cardverificationcodesupplied` — cvc code supplied\n- `cvcresponsecode` — response code from the issuer concerning the cvc code\n-  `card_id` — a hashed version of the card number",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The evaluation metric for this competition is [area under the ROC curve ](https://en.wikipedia.org/wiki/Receiver_operating_characteristic) between the predicted probability and the observed target.\n\n## Submission Format\nFor each `txid` in the test set, you must predict the probability if it is a fraudulent transaction or not. The file should contain a header and have the following format:\n\n```\ntxid, is_fraud\n0, 0.5\n1, 0.5\n```",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}