Herded Stein Weighted#
Example coreset generation using randomly generated point clouds.
This example showcases how a coreset can be generated from a dataset containing n
points sampled from k clusters in space.
A coreset is generated using Stein kernel herding, with a PCIMQ base kernel.
The initial coreset generated from this procedure is then weighted, with weights determined such that the weighted coreset achieves a better maximum mean discrepancy when compared to the original dataset than the unweighted coreset.
The coreset attained from Stein kernel herding is compared to a coreset generated via uniform random sampling. Coreset quality is measured using maximum mean discrepancy (MMD).
- examples.herding_stein_weighted.main(out_path=None)[source]#
Run the tabular herding example using weighted herding.
Generate a set of points from distinct clusters in a plane. Generate a coreset via weighted herding. Compare results to coresets generated via uniform random sampling. Coreset quality is measured using maximum mean discrepancy (MMD).