Code for the paper "An Online Algorithm for Learning Buyer Behavior under Realistic Pricing Restrictions"
This codebase is for algorithms proposed in the aforementioned paper. It consists of four python scripts:
- algorithms.py : has the proposed and competitor algorithms
- data.py : generates data (e.g., from the billion prices project)
- experiments.py : (calls algorithms on buyer models)
- buyers.py : has the four buyer models
- geometric.py : has definitions for primitives such as ellipsoids and halfspaces that are used in the algorithms.
The easiest way to get started is to look at experiments.py and go from there.
Some useful pointers:
To reload Ipython modules (useful for debugging) following this:
%load_ext autoreload %autoreload 2
This can be saved in
c.InteractiveShellApp.extensions = ['autoreload'] c.InteractiveShellApp.exec_lines = ['%autoreload 2']