From the prudsys task description:
###Scenario The central topic of this year’s task in the DATA MINING CUP competition is the prediction of the return rates for a fashion distributor. Returns have been a major cost driver for online shops for many years. This is particularly the case for assortments in the fashion industry. Many approaches to solve this problem are based on forecasting models. Predicting returns was the central topic in the DATA MINING CUP of 2014. This year we look at the topic in more detail focusing in particular on the influence of discounted items and vouchers on return rates.
###About the task A fashion distributor sells articles of particular sizes and colors to its customers. In some cases items are returned to the distributor for various reasons. The order data and the related return data were recorded over a two-year period. The aim is to use this data and methods of data mining to build a model which enables a good prediction of return rates.