Skip to content

ekapope/Banking-Data-Analysis-Reveals-Non-Performing-Customers

Repository files navigation

Datamart Report - PKDD'99 Discovery Challenge

The purpose of this report is to analyze the financial status of clients in Czech Republic between January 1st ,1993 and December 31st, 1998. In order to create data mart, we used the following datasets 1.Card 2.Client 3. Account 4.Trans 5. Orders 6.Loans 7.Demographics 8. Disp. The total observations and variables of this data mart are 5,369 and 81 respectively

Output: Datamart Analysis - Reveals Top & Bottom 10% customers

‘trans_ratio_98_97’ column derives from the ratio of transaction growth in the past year. After sorted this column descending, it reveals the top 10% customers with the most transaction growth in the past year and also the bottom 10% which had the least transaction growth. With this information, the bank can take appropriate action on this.

The client data was exported into two csv files; bottom10percent_customer.csv, top10percent_customer.csv.

Data source:

PKDD 1999, Discovery Challenge Guide to the Financial Data Set (1999)

About

Transaction data analysis using Python.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages