Order receipt of bank transactions

This type of data is generated by financial institutions and is usually protected and not made available to the public.

There is a dataset, that has real transactions but it doesn't have any label for fraud detection. It is Berka dataset available as part of PKDD'99 Discovery Challenge. It is real anonymized data from the Czech bank.

• http://sorry.vse.cz/~berka/challenge/pkdd1999/data_berka.zip
• http://lisp.vse.cz/pkdd99/berka.htm

It seems the link for data description doesn't work anymore, so please use this link from WebArchive:

• https://web.archive.org/web/20161019192412/http://lisp.vse.cz/pkdd99/berka.htm

Python Notebooks that did some analysis on the data that might be helpful:

• https://github.com/justinng1/berka

Follow this thread

---

Santander Customer Transaction Prediction

You can find more at Kaggle Datasets