Skip to content

Releases: DoktorMike/EvidentialFlux.jl

v1.6.0

09 Nov 19:54

Choose a tag to compare

1.6.0 (2025-11-09)

Features

  • added new losses for regression. (b0805be)
  • added the uncertainty corrected loss for DER. (326ca81)

Documentation

  • 📛 badges addition for new formatter. (82aec5c)

v1.5.3

30 Sep 20:36

Choose a tag to compare

1.5.3 (2025-09-30)

Bug Fixes

  • finally fixed the damn documentation. (a23251c)

Documentation

  • clarified the predict function for MVE networks. (4973a4c)

Other

  • 🦋 formatting according to runic. (8e8888c)

v1.5.2

19 Jun 16:05

Choose a tag to compare

1.5.2 (2024-06-19)

Bug Fixes

  • fixed a bug in the prediction function of MVE network. (fcada7d)

v1.5.1

18 Jun 12:14

Choose a tag to compare

1.5.1 (2024-06-18)

Features

  • Made explicit mu and sigma layers for MVE network. (97165ab)

Documentation

v1.5.0

18 Jun 12:13

Choose a tag to compare

v1.5.0 Pre-release
Pre-release

1.5.0 (2024-06-18)

Features

  • Made explicit mu and sigma layers for MVE network. (97165ab)

Documentation

v1.4.0

17 Jun 11:40

Choose a tag to compare

1.4.0 (2024-06-17)

Features

  • Implemented Mean-variance network and cleaned up the dependencies. (448e865)

v1.3.3

17 Jun 11:39

Choose a tag to compare

1.3.3 (2023-04-25)

Other

  • Added automatic formatting check for SciML compliance. (3fe21d8)

Documentation

  • Added a badge for the style compliance check. (8bdbee8)
  • Added a DOI Zenodo badge. (ea3f035)
  • small corrections (708f4b1)

v1.3.2

26 Jul 20:09

Choose a tag to compare

Full Changelog: v1.3.1...v1.3.2

v1.3.1

21 Jul 20:28

Choose a tag to compare

What's Changed

Full Changelog: v1.3.0...v1.3.1

v1.3.0

19 Jul 21:20

Choose a tag to compare

1.3.0 (2022-07-19)

Features

  • Implemented the DER correction 🥶 (450a76f), closes #9
  • Updated regression example to produce something more useful and more inline with the paper. (8577d8e)

Documentation

  • Added classification image to readme. (9b3f64f)
  • Added documentation of the dirloss function. (e412f92)
  • Added more text regarding the classification entry. (afb03d1)
  • Added the documentation of the DIR layer. (736ea62)
  • Added the regression case to the README. (5fb3060)
  • Documented the Deep evidential classification. (f33f872)
  • Updated the example for the second regression case to use a power of 2 in the loss. (ec2026b)