This is a program, in python 3.5, which is able to learn acyclic conditional preference networks from noisy preferences.
It provide two algorithms : a batch algorithm and an online algorithm based on the McDiarmid bound in order to take the best decision at the best moment.
You can use the test files in the project root to use the algorithms. You can use your own database, please follow the format in the datasets. You also can generate a random database with noise.
I am sorry for this short readme for the moment, it correspond to my actual work as a Ph.D student in computer science. More informations will be available soon!