2 Data

The data set contains data on all 191’779 members of the organization with a lapsed donation status (last donation 13 – 24 months ago) relative to the promotion sent out in June 1997 (named current promotion hereafter).

The data is provided³ split in two sets, of which one is intended for learning (95’412 examples), the other for validation and final prediction (96’367 examples). The features are identical between the two except for the target features that have been separated from the validation set.

In this section, the learning data set will be characterized.

---

3. See https://archive.ics.uci.edu/ml/datasets/KDD+Cup+1998+Data, accessed on 30.5.2019↩