Coreax – Revolutionizing Data Reduction With Coreset Algorithms In JAX

For n points in d dimensions, a coreset algorithm takes an n×d data set and reduces it to m≪n points whilst attempting to preserve the statistical properties of the full data set.

The algorithm maintains the dimension of the original data set. Thus the m points, referred to as the coreset, are also d-dimensional.

The m points need not be in the original data set. We refer to the special case where all selected points are in the original data set as a coresubset.

Some algorithms return the m points with weights, so that importance can be attributed to each point in the coreset. The weights, wi for i=1,…,m, are often chosen from the simplex. In this case, they are non-negative and sum to 1: wi>0 ∀i and ∑iwi=1.

Please see the documentation for some in-depth examples.

Example Applications

Choosing Pixels From An Image

In the example below, we reduce the original 180×215 pixel image (38,700 pixels in total) to a coreset approximately 20% of this size. (Left) original image.

(Centre) 8,000 coreset points chosen using Stein kernel herding, with point size a function of weight. (Right) 8,000 points chosen randomly. Run examples/david_map_reduce_weighted.py to replicate.

Setup

Before installing coreax, make sure JAX is installed. Be sure to install the preferred version of JAX for your system.

Install JAX noting that there are (currently) different setup paths for CPU and GPU use:

$ python3 -m pip install jax

For more information click here.