Latest commit d3075b8 Jan 22, 2015 Olivier Chapelle New demo using conversions logs from Criteo
Permalink
..
Failed to load latest commit information.
naive_baseline.py
readme.txt

readme.txt

        ------ Criteo Conversion Logs ------

This dataset contains feature values and conversion feedback for millions of clicks on display ads.
It served to benchmark algorithms described in the paper "Modeling Delayed Feedback in Display Advertising", KDD 2014 by O.Chapelle (http://olivier.chapelle.cc/pub/delayedConv.pdf)

===================================================

Instructions

 - Go to http://labs.criteo.com/downloads/2014-conversion-logs-dataset/ 
 - Download the tar file after agreeing to the terms and conditions
 - Untar the file 
 - Move data.txt to this directory
 - Run the naive_baseline.py script

====================================================

Dataset assembled by Olivier Chapelle (o.chapelle@criteo.com)