Submodularity in Data Science
Mohammad Reza Karimi
Data Science Summer School, June 2018, Paris
Welcome to the tutorial on Submodularity. To begin, please download this zip file containing notebooks, datasets, and the submodpy package. The archive can be obtained from here.
To be able to run the codes presented in these notebooks, you need to install the following packages:
- jupyter, ipywidgets
- numpy, scipy, matplotlib, scikit-learn, scikit-image
- networkx, nxpd (for plotting graphs)
- cvxopt, cvxpy
- pycddlib, pypoman (optional; can be obtained from here and installed manually)
You can install the required packages via pip:
pip install -r requirements.txt
Note. The submod package offered to you is just meant to be used in the summer school and is a small subset of a bigger package which is going to be published soon. Almost all of the code in this package is written by Josip Djolonga.