Transaction data possessed by banks contain volumes of information about the bank’s customers and their financial behaviour. Analysing such data could help gain interesting insights, and find new trends or patterns in customer behaviour. However, for privacy and security reasons, such data is kept undisclosed.

The transaction dataset generated here is a synthetic dataset which contains fund transfer information of a few million accounts. It contains about 5 million transactions all spanning over a time period of 1 year (2017), and contains essential information such as amount transferred/deposited/withdrawn, date and time of transfer (in unix timestamp format), type of transfer, etc. (The Dataset Schema has detailed information about the same).

Usage examples

  1.  Analysis of the dataset for suspicious financial activity.
  2.  Analysis of customer data to find patterns in their financial behaviour.
  3.  Predicting future behaviour of customers by analysing current trends.

Download Dataset Schema

Your feedback on the quality of the dataset will be very helpful, please fill the form and press the Download button to download the dataset!

Please mention the assistance you received from this dataset in any relevant material you publish. You may add a citation text as follows:

Patil, Karan (2018), Synthetic Dataset: Banking Transactions Dataset, aCubeIT Machine Learning Dataset Repository [].

Contributed by: Karan Patil, ML Engineer, aCubeIT