Approximate hubness reduction

The Hub Toolbox now supports approximate hubness reduction for very large data sets. Please check out the approximate module for hubness reduction with linear time and space complexity.

v2.3.1: Hub Toolbox 2.3.1

This is primarily a maintenance release.

- Increased unit test coverage
- documentation
- bug fixes

Hub Toolbox 2.3.0

The HUB TOOLBOX now supports the following hubness reduction methods:

• Mutual Proximity (Empiric, Gauss, Indep. Gauss, Indep. Gamma)
• Local Scaling (Normal, NICDM)
• Shared Nearest Neighbors
• Centering (Normal, Weighted, Localized)
• DisSim (Global, Local)

Most methods now support (sparse) similarity matrices.

Most methods now support train/test splits.

MP and hubness functions now support parallel processing.

Performance improvements for several methods.

Vast unit test coverage.

Hub Toolbox 2.2

Features (experimental) centering functions for hubness reduction.

Hub Toolbox 2.1

v2.1.0

minor changes