Density estimation in kernel exponential families.
dekef is a python package providing tools for
- computing density estimates in kernel exponential families using the score matching loss function and the negative log-likelihood loss function,
- visualizing the density estimates when data are 1-dimensional or the contour plot of the density estimates when data are 2-dimensional, and
- assessing the quality of density estimates using the correlation and the mean integrated squared error (MISE).
Please refer to Jupyter notebooks in /examples for examples of
how to use functions to compute and visualize density estimates
and assess their qualities.
Chenxi Zhou – chenxizhou.jayden@gmail.com