CHANGELOG.rst

Changelog

Badge	Meaning
<span class="badge badge-success">Feature</span> {\small\sc [Feature]}	Add something that cannot be achieved before.
<span class="badge badge-info">Efficiency</span> {\small\sc [Efficiency]}	Improve the efficiency on the computation or memory.
<span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]}	Miscellaneous minor improvements.
<span class="badge badge-danger">Fix</span> {\small\sc [Fix]}	Fix up something that does not work as expected.
<span class="badge badge-warning">API Change</span> {\small\sc [API Change]}	You will need to change the code to have the same effect.

Ver 0.1.*

• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix the issue that model will be overwritten when using the validation data loder | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} Add internal unsqueeze operation in forward of all classifiers | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add voting_strategy parameter for VotingClassifer, NeuralForestClassifier, and SnapshotEnsembleClassifier | @LukasGardberg
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix the sampling issue in BaggingClassifier and BaggingRegressor | @SunHaozhe
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add NeuralForestClassifier and NeuralForestRegressor | @xuyxu
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Relax check on input dataloader | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Support arbitrary training criteria for all ensembles except Gradient Boosting | @by256 and @xuyxu
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix missing functionality of save_model for fit of Soft Gradient Boosting | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add SoftGradientBoostingClassifier and SoftGradientBoostingRegressor | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Support using dataloader with multiple input | @xuyxu
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix missing functionality of use_reduction_sum for fit of Gradient Boosting | @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Relax tensorboard as a soft dependency | @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Simplify the training workflow of FastGeometricClassifier and FastGeometricRegressor | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Support TensorBoard logging in set_logger | @zzzzwj
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add use_reduction_sum parameter for fit of Gradient Boosting | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Improve the functionality of evaluate and predict | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add FastGeometricClassifier and FastGeometricRegressor | @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Add flexible instantiation of optimizers and schedulers | @cspsampedro
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add support on accepting instantiated base estimators as valid input | @xuyxu
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix missing base estimators when calling load() for all ensembles | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add methods on model deserialization load() for all ensembles | @mttgdd

Beta

• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add set_scheduler for all ensembles | @xuyxu
• <span class="badge badge-success">Major Feature</span> {\small\sc [Major Feature]} Add AdversarialTrainingClassifier and AdversarialTrainingRegressor | @xuyxu
• <span class="badge badge-success">Major Feature</span> {\small\sc [Major Feature]} Add SnapshotEnsembleClassifier and SnapshotEnsembleRegressor | @xuyxu
• <span class="badge badge-success">Feature</span> {\small\sc [Feature]} <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Add model validation and serialization | @ozanpkr and @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Add CI and maintenance tools | @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Add the code coverage on codecov | @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Add the version numbers to requirements.txt | @zackhardtoname and @xuyxu
• <span class="badge badge-primary">Enhancement</span> {\small\sc [Enhancement]} Improve the logging module using logging | @zzzzwj
• <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Remove the input argument output_dim from all methods | @xuyxu
• <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Refactor the setup on optimizer into set_optimizer | @xuyxu
• <span class="badge badge-warning">API Change</span> {\small\sc [API Change]} Refactor the codes on operating tensors into an independent module | @zzzzwj
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix the bug in logging module when using multi-processing | @zzzzwj
• <span class="badge badge-danger">Fix</span> {\small\sc [Fix]} Fix the binding problem on scheduler and optimizer when using parallelization | @Alex-Medium and @xuyxu