v2.3.2

What's Changed

• fix: force UTF-8 encoding on HTTP response body + CLI docs by @cardmagic in #131

Full Changelog: v2.3.1...v2.3.2

v2.3.1

What's Changed

• fix: force UTF-8 encoding for non-UTF8 locales by @cardmagic in #130

Full Changelog: v2.3.0...v2.3.1

v2.3.0

What's Changed

• docs: fix README examples and broken links by @cardmagic in #113
• ci: add Dependabot for automated dependency updates by @cardmagic in #115
• chore(deps): bump matrix from 0.4.2 to 0.4.3 by @dependabot[bot] in #117
• chore(deps-dev): bump rdoc from 6.5.1.1 to 7.0.3 by @dependabot[bot] in #116
• Add CLI executable with model registry support by @cardmagic in #127

Full Changelog: v2.2.0...v2.3.0

v2.2.0

What's Changed

• feat(bayes): add keyword argument API for train and untrain by @cardmagic in #97
• fix(lsi): improve sentence and paragraph splitting in Summary by @cardmagic in #98
• Add property-based tests for probabilistic invariants by @cardmagic in #99
• feat(lsi): add hash-style API for adding items by @cardmagic in #101
• feat(knn): add k-Nearest Neighbors classifier by @cardmagic in #105
• feat(tfidf): add TF-IDF vectorizer by @cardmagic in #107
• feat: Add Logistic Regression classifier by @cardmagic in #108
• feat: Add streaming training and incremental SVD support by @cardmagic in #109
• fix: accept array of categories in classifier initialization by @cardmagic in #111
• Release v2.2.0 by @cardmagic in #112

Full Changelog: v2.1.0...v2.2.0

v2.1.0

What's Changed

• Add zero-dependency native C extension for LSI acceleration by @cardmagic in #89
• feat: add thread safety to Bayes and LSI classifiers by @cardmagic in #86
• feat: add save/load methods for classifier persistence by @cardmagic in #85
• perf: cache expensive computations in Bayes classifier by @cardmagic in #84
• feat(lsi): expose tuning parameters with validation and introspection API by @cardmagic in #92
• Add rubyclassifier.com links to README and gemspec by @cardmagic in #94
• feat: add pluggable persistence backends with storage API by @cardmagic in #93

Full Changelog: v2.0.0...v2.1.0

v2.0.0

What's Changed

• Fix CI: use supported Ruby versions by @cardmagic in #44
• Fix: LoadError by @maxhungry in #42
• Refactor nested conditionals to early returns by @cardmagic in #43
• Fix Gemfile.lock to include matrix dependency by @cardmagic in #47
• Add RBS type signatures by @cardmagic in #49
• Audit spec coverage and add SimpleCov by @cardmagic in #53
• Add tests for Bayes#untrain method by @cardmagic in #55
• Add edge case tests for text handling by @cardmagic in #58
• Add test for remove_item triggering needs_rebuild by @cardmagic in #59
• Add RuboCop for code standardization by @cardmagic in #61
• Fix numerical stability issues in SVD implementation by @cardmagic in #73
• Add LSI benchmark system for GSL vs native Ruby comparison by @cardmagic in #74
• Modernize README to follow Ruby open source best practices by @cardmagic in #77
• Implement proper Laplace smoothing in Bayes classifier by @cardmagic in #80
• Fix LSI dimension mismatch with native Ruby SVD by @cardmagic in #78
• Add GitHub Actions workflow for automated gem releases by @cardmagic in #81
• Use .sum instead of .inject for summing by @cardmagic in #82
• Migrate from separate RBS files to inline annotations by @cardmagic in #83

New Contributors

• @maxhungry made their first contribution in #42

Full Changelog: v1.4.4...v2.0.0

v1.4.4

Release Notes for Classifier 1.4.4

New Features and Enhancements

1. Fix bug in classifying text with repeated words
   • There was a bug classifying LSI text with repeated words that is now fixed

v1.4.3

Release Notes for Classifier 1.4.3

New Features and Enhancements

1. Fix support for Ruby 2.7
   • Make sure that Ruby 2.7 works

v1.4.2

Release Notes for Classifier 1.4.2

New Features and Enhancements

1. Fix bug in remove_category Method

   • Fixed a race condition mentioned by @dpetruha here 17143ab

2. Add more unit tests

   • Added unit tests for the remove_category feature

v1.4.1

Release Notes for Classifier 1.4.1

New Features and Enhancements

1. Add remove_category Method

   • A new remove_category method has been added to the Bayes class. The changes aim to provide users with more control over the Bayesian classifier by allowing them to remove categories. This feature can be useful in scenarios where categories become obsolete or need to be restructured. PR: #39 Closes: #12

2. Add more unit tests

   • Added unit tests for the remove_category feature