Releases: remisultan/java-ml
Releases · remisultan/java-ml
1.2.0
What's New !
- Principal Component Analysis
- Better CSV parsing with Univocity
- Train Test Shuffle Dataframes
- Decision Tree Classifier / Regressor
- Random Forest Classifier / Regressor
1.1.1
Fixes
- Linear regression data normalizing done wrongly
Misc
- Better logging for KMeans / Kmedians
1.1.0
What's New !
- double bandwidth for MedoidShift
- better medoid shift logging
FIxes
- format negative numbers
Misc
- more complete README docs
1.0.1
Changelog
Fixed
- corrected groupId and artifactId
1.0.0
Changelog
What's new
- logistic regression
- softmax regression
- kmedoids clustering
- mean and median shift
- kmedoid evaluator
- plus plus strategy for Kmedoid
- dataframe filter
- regularization of loss and gradient
- implement multivariate linear regressionand packages
Fixes
- using a functional wagon storage version
- permission on file execution
- right shape of matrix when empty
- argument not being initialized properly + docs on dataset
Misc
- refactor: kmedoids refactor with image example
- test: changing test due to normalization with mean
- refactor: naming things properly + fix on dataframe printer
- refactor: interfaces ar regression