zhanpython / gp_pref_elicit

Ordered Preference Elicitation Strategies for Multi-Objective Decision Making using Gaussian Processes

This is the code for the paper

Ordered Preference Elicitation Strategies for Supporting Multi-Objective Decision Making
Luisa M Zintgraf, Diederik M Roijers, Sjoerd Linders, Catholijn M Jonker, Ann Nowe
17th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2018)

which you can find on arXiv.

Main Implementation

The main code, the implementation of pairwise gaussian processes, is comprised of gaussian_process.py, acquisition_function.py, and dataset.py.

For examples on how to use them, we refer to gp_utilities/utils_experiment.py and to the experiments folder. There you can also find all experiments we ran for the paper. If you want to run something and get pretty plots, try

python experiments/exp_gp-shape.py

Web Interface

The source code for the user study can be found in the folder webInterface. To run it, type

python webInterface/start.py

in a terminal and go to http://0.0.0.0:5000/ in a web browser of your choice (I tested a few, but can't guarantee it works everywhere). There you should find the starting page of the experiment. Flags: -t to skip the tutorial, -d for debug mode.