Efficient and principled score estimation with Nyström kernel exponential families
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
notebooks
nystrom_kexpfam
scripts
.gitignore
LICENSE
README.md

README.md

Experimental codes for AISTATS 2018 paper "Efficient and principled score estimation with Nyström kernel exponential families" by Dougal Sutherland, Heiko Strathmann, Michael Arbel, and Arthur Gretton, https://arxiv.org/abs/1705.08360.

See notebooks/demo.ipynb for how to use the estimator(s), and how to replicate experimental results.

Dependencies (some are optional, see demo notebook):

For the Python packages (given that you have downloaded them) and Shogun (given that you have compiled or installed it), this could be achieved with something like

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:path/to/libshogun.so
export PYTHONPATH=$PYTHONPATH:/path/to/shogun.py
export PYTHONPATH=$PYTHONPATH:/path/to/nystrom-kexpfam
export PYTHONPATH=$PYTHONPATH:/path/to/kernel_exp_family