PAC-Bayesian Domain Adaptation (aka PBDA) -- machine learning algorithm
Clone or download
GERMAIN Pascal
Latest commit b436313 Oct 30, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.gitignore Initial commit May 31, 2013
LICENSE First commit of the code. May 31, 2013
README Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018
__init__.py
common.py Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018
dataset.py
kernel.py Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018
pbda.py
pbda_classify.py Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018
pbda_learn.py Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018
pbda_reverse_cv.py Update to Python 3 and a bit of refreshing, because we are (almost) i… Oct 30, 2018

README

----------------------------------------------------------------------------------------------------
PAC-BAYESIAN DOMAIN ADAPTATION (aka PBDA)
Version 0.901 (August 9, 2013), Released under the BSD-license
https://github.com/pgermain/pbda 
----------------------------------------------------------------------------------------------------
Author: 
    Pascal Germain. Groupe de Recherche en Apprentissage Automatique de 
    l'Universite Laval (GRAAL).

Reference: 
    Pascal Germain, Amaury Habrard, Francois Laviolette, and Emilie Morvant. 
    A PAC-Bayesian Approach for Domain Adaptation with Specialization to 
    Linear Classifiers.
    International Conference on Machine Learning (ICML) 2013. 
---------------------------------------------------------------------------------------------------- 

Thank you for looking at my code!

This program have been tested using Python 3.6 under Linux.
It requires the NumPy and SciPy libraries.

I prepared three small scripts to use PBDA by the command line:
1) pbda_learn.py: Execute the learning algorithm
2) pbda_classify.py: Execute the classification function
3) pbda_reverse_cv.py: Compute a "reverse cross-validation" score

Further usage instructions can be obtained by the following commands:
python pbda_learn.py --help
python pbda_classify.py --help
python pbda_reverse_cv.py --help

The data used in the paper experiments is available here (in svmlight format):
http://researchers.lille.inria.fr/pgermain/data/amazon_tfidf_svmlight.tgz

Pascal Germain.