Features

The 13th Pacific-Asia Knowledge Discovery and Data Mining conference (PAKDD 2009) is pleased to host another PAKDD 2009 data mining competition, co-organized by NeuroTech Ltd. and Center for Informatics of the Federal University of Pernambuco (Brazil).

Competitions in scientific events have been organized world-wide for stimulating the application of state-of-the-art approaches to real world problems. In recent years, PAKDD has organized several data mining competitions and this year presents a problem on the well known application of credit scoring. The main novelty is the LeaderBoard for stimulating the competitors� participation by assessing and ranking their preliminary solutions on an unofficial data set.

The competition is open for academia and industry. The only ineligible participants are staff and students from Center for Informatics of the Federal University of Pernambuco and NeuroTech Ltd.

Problem Summary

Credit Risk Assessment on a Private Label Credit Card Application

The offer of credit for potential clients is a very important service for stimulating consumption in the market. Despite being among the oldest application domains for data mining (from the times when not even the name "data mining" existed), there are some difficulties related to credit scoring which are often overlooked by modelers, namely:

• In general, there are only data about the company's clients for modeling, but not about the rejected applicants. These represent a sample of the potential clients (market) that is strongly biased given that a systematic procedure focused on the problem target (payment default) has been applied for selection.
• Also, the data has been collected from a time interval in the past for developing a model to be applied in a future time. Not considering any drastic change in the economy, gradual market changes occur and reduce the performance of the model estimated on the modeling data set.

This competition focuses on the credit scoring model's robustness against performance degradation caused by market gradual changes along few years of business operation.

Participants will download a labeled data set from one year period for modeling, will submit scores to the leaderboard dataset from one year later, and the competition results will be assessed on scores submitted for a prediction dataset starting from three years later. These data sets come from the private label credit card operation of a major Brazilian retail chain, along stable inflation condition (2003-2008).

The official performance metrics will be the area under the ROC curve and a Java routine for calculating it is available for download.

Important Dates

Feb 12 - Competition page open

• Modeling data set release
• Competition announcement
  Feb 16 - LeaderBoard data set release
• LeaderBoard open for submissions
  Mar 02 - Prediction data set release
• Prediction submission open (for manuscript and scores)
  Mar 20 - Competition submission deadline (PDF manuscript and scores)
  Apr 03 - Competition results released
  Apr 27 - Conference starts