Skip to content

criteo-research/criteo-privacy-preserving-ml-competition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Criteo Privacy-preserving Machine Learning Competition @ AdKDD'21

In the Criteo-AdKdd'21 competition, the goal was to learn a ML model from aggregated data.

This repository contains useful code and documentation for the challenge, and the notebooks to run the experiments reported in the research article.

Related Works

RELATEDWORKS.md contains a short selection of links to potentially interesting works relating Privacy and ML. Please feel free to submit pull requests to enhance this selection.

Fetch the data

In the data directory you can find two notebooks:

  • fetch_datasets: allows to retieve all the competition data as well as the granular data used to make the aggregations and another larger test set, used in our experiments to evaluate the performance of the models according to the number of samples.
  • generate_aggregated_datasets: notebook that we used to generate the aggregations. It is necessary to run it to create the noiseless aggregated datasets that we use in the experiments to study the robustness to noise of the different methods.

Running experiments

Gradient Boosting Trees

Models and experiments performed with Gradient Boosting Trees can be found in the gbt folder. There are two notebooks, one for clicks and one for sales.

Logistic Regression

Models and experiments performed with a Logistic Regression can be found in the logistic_regression folder in the main notebook. encode_data notebook should be run first to one-hot encode in a common space the modalities of each feature and the pairs of modalities of each pair of feature.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published