Data

Tools

Indexes

Applications

Experiments

Open Source Science

Recent Releases of scikit-activeml

scikit-activeml - 0.6.2

We are happy to announce the following changes/features:

Installation

  • Upload to PyPI complies with PEP625 by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/499

Tests

  • Improve formatting and reproducibility by @mherde and @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/505

Requirements

  • Update joblib requirement from <=1.4.2 to <=1.5.0 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/497
  • Update makefun requirement from <=1.15.6 to <=1.16.0 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/500

Full Changelog: https://github.com/scikit-activeml/scikit-activeml/compare/0.6.1...0.6.2

- Python
Published by tpham93 10 months ago

scikit-activeml - 0.6.1

We are happy to announce the following changes/features:

Bugfixes

  • #494 Add fix for repeated fit for ParzenWindowClassifier with mean criterion by @tpham93.
  • #493 Fix RT-AL for method="representativity" by @mherde. Thanks to @bjaster for pointing out this bug and proposing a fix in #491.

Full Changelog: 0.6.0...0.6.1

- Python
Published by tpham93 10 months ago

scikit-activeml - 0.6.0

We are happy to announce the following changes/features:

Tutorials

Query Strategies

Estimators

Documentation

Continuous Integration

Requirements

Other

  • #481 Fix bug in match_signature with type hints by @tpham93

References

  • [1] S. Gilhuber, A. Beer, Y. Ma, and T. Seidl. FALCUN: A Simple and Efficient Deep Active Learning Strategy. In Jt. Eur. Conf. Mach. Learn. Knowl. Discov. Databases, 421–439. 2024.
  • [2] S. R. Gupte, J. Aklilu, J. J. Nirschl, and S. Yeung-Levy. Revisiting Active Learning in the Era of Vision Foundation Models. Trans. Mach. Learn. Res., 2024.
  • [3] V. C. Raykar, S. Yu, L. H. Zhao, G. H. Valadez, C. Florin, L. Bogoni, and L. Moy. Learning from Crowds. J. Mach. Learn. Res., 11(4):1297–1322, 2010.

Full Changelog: 0.5.2...0.6.0

- Python
Published by mherde 11 months ago

scikit-activeml - scikit-activeml 0.5.2

We are happy to announce the following changes/features:

Query Strategies:

  • Fix TypiClust when batch_size is higher than the number of non-empty clusters by @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/452

Documentation:

  • Add development branch to documentation by @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/448

Continuous Integration:

  • Add workflow parallelization by @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/441

Requirements:

  • Update scipy requirement from <=1.14.0 to <=1.14.1 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/449
  • Update numpy requirement from <=2.0.1 to <=2.1.0 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/445
  • Update matplotlib requirement from <=3.9.1 to <=3.9.2 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/446

Full Changelog: https://github.com/scikit-activeml/scikit-activeml/compare/0.5.1...0.5.2

- Python
Published by mherde over 1 year ago

scikit-activeml - scikit-activeml 0.5.1

We are happy to announce the following changes/features:

Query Strategies:

  • Implement pool-based query strategy Probability Coverage (ProbCover) [1] as ProbCover by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/426
  • Implement pool-based query strategy Contrastive Active Learning (CAL) [2] as ContrastiveAL by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/430
  • Implement pool-based query strategy Clustering Uncertainty-weighted Embeddings (CLUE) [3] as Clue by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/437
  • Add sampling option and variation ratios [4] as new options to pool-based query strategy Query by Committee (QBC) in QueryByCommittee by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/434

Tutorials:

  • Add "Open in Colab" button to all tutorial notebooks and examples by @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/429

Documentation:

  • Implement versioning of the documentation and parallelize examples by @AlexanderBenz in https://github.com/scikit-activeml/scikit-activeml/pull/379 and by @tpham93 in https://github.com/scikit-activeml/scikit-activeml/pull/438
  • Add community standards by @mherde in https://github.com/scikit-activeml/scikit-activeml/pull/431

Requirements:

  • Update optional maximum numpy requirement from <=2.0.0 to <=2.0.1 by @dependabot in https://github.com/scikit-activeml/scikit-activeml/pull/425

References

Full Changelog:

https://github.com/scikit-activeml/scikit-activeml/compare/0.5.0...0.5.1

- Python
Published by mherde over 1 year ago

scikit-activeml - scikit-activeml 0.5.0

We are happy to announce the following changes/features:

Wrapper Query Strategies: - FR: Addition of the pool-based query strategy wrapper SubSamplingWrapper, which allows subsampling a smaller set of sample candidates to accelerate the sample selection via query (cf. #349). - FR: Addition of the pool-based query strategy wrapper ParallelUtilityEstimationWrapper, which allows parallelizing the utility computation across the sample candidates via joblib. As a requirement, the utility computation must be independent between the candidate samples (cf. #402).

Query Strategies: - FR: Addition of the pool-based query strategy RegressionTreeBasedAL, which leverages a regression tree to select samples for regression tasks (cf. #371). - FR: Addition of the pool-based query strategy CoreSet, which leverages kmeans++ to select diverse samples. Originally, this strategy has been designed for classification tasks. Yet, it can also be used in the context of regression (cf. #281). - FR: Addition of the pool-based query strategy TypiClust, which leverages kmeans to balance the trade-off between diversity and representativeness (typicality) of samples. Originally, this strategy has been designed for classification tasks. Yet, it can also be used in the context of regression (cf. #364). - FR: Addition of the pool-based query strategy Badge, which leverages kmeans++ in the gradient space of the last linear layer of a classification neural network. Thereby, its sample selection balances the trade-off between diverse and uncertain samples. - FR: Addition of the pool-based query strategy GreedyBALD, which performs a top-k selection for the utilities computed according to BALD (cf. #323). - BUG: Fix the issue of estimators outputting zero probabilities, raising warnings when computing log probabilities in BatchBALD. - BUG: Fix the handling of candidates in the multi-annotator pool-based query strategy wrapper SingleAnnotatorWrapper (cf. #404) and ensure that missing_label the wrapper and the wrapped single-annotator query strategy match (cf. #405).

Estimators: - FR: SklearnClassifier, SklearnClassifier, and SlidingWindowClassifier have match_signature as new decorators for the methods to enforce the correct mapping between the parameter signature between the wrapper's methods and the wrapped estimator's methods (cf. #375).

Tutorials: - FR: Add a tutorial showcasing query strategies suited for deep learning tasks in combination with self-supervised models to extract self-supervised sample features (cf. #282).

Requirements: - FR: Update of supported Python versions to 3.9, 3.10, 3.11, and 3.12 (cf. #378, #416). The requirements are updated accordingly. - FR: Removed upper bounds in requirements to avoid undesired downgrades of packages. Still, a guaranteed working installation can be enforced via max_requirements.txt. - FR: Add makefun as a new requirement to match signatures when using wrappers (cf. #375). - FR: Implement automatic dependency check via dependabot (cf. #302, #314).

Tests: - Add and use new test templates for classifiers, regressors, stream-based query strategies, and budget managers (cf. #303).

Documentation: - FR: Add note that outputs of the query strategy examples are given at the bottom of each example (cf. #304).

Continuous Integration: - FR: Addition of automatic formatting check via black (cf. #301). - FR: Parallelize example generation to speed up the CI process (cf. #306).

Full Changelog: - https://github.com/scikit-activeml/scikit-activeml/compare/0.4.1...0.5.0

- Python
Published by mherde over 1 year ago

scikit-activeml - scikit-activeml 0.4.1

We are happy to announce the following changes/features:

  • We added CogDQS and DBALStream as two new stream-based query strategies (cf. #254).
  • We implemented BatchBALD as a new pool-based query strategy (cf. #284).
  • We added examples for the stream-based query strategies to the documentation and updated the notion of utilities, which now need to be maximized (cf. #286).
  • We added legends to the examples for better understanding (cf. #287).
  • We revised the developer guide (cf. #290).

- Python
Published by mherde almost 3 years ago

scikit-activeml - scikit-activeml 0.4.0

We are happy to announce the following changes/features:

  • We now support Python 3.10, and we deprecated Python 3.7 (cf. #279).
  • We now support scikit-learn 1.2 (cf. #279).
  • Update of probabilistic active learning to support mean bandwidth kernel (cf. #260).
  • Added examples for DWUS and DUAL as two newly supported uncertainty-based query strategies (cf. #266).
  • Updated QUIRE to allow sampling when mathematical assumptions are not fulfilled (cf. #274 ).

- Python
Published by mherde about 3 years ago

scikit-activeml - scikit-activeml 0.3.1

We are happy to announce the following changes/features:

  • Improved test infrastructure
  • Documentation improved
  • Test coverage does not include test files anymore

- Python
Published by dakot over 3 years ago

scikit-activeml - scikit-activeml 0.3.0

We are happy to announce the following changes/features:

  • Added pool-based query strategies for regression
  • Fixed bug in vote entropy for query-by-committee
  • Improved documentation

- Python
Published by dakot over 3 years ago

scikit-activeml - scikit-activeml 0.2.5

We are happy to announce the following changes/features:

  • Added an alternative kernel frequency estimation option for StreamProbabilisticAL

- Python
Published by dakot over 3 years ago

scikit-activeml - scikit-activeml 0.2.4

We are happy to announce the following changes/features:

  • Implementation of Discriminative Active Learning as a new pool-based query strategy
  • Implementation of a new classifier wrapper to enable learning with sliding windows

- Python
Published by tpham93 over 3 years ago

scikit-activeml - scikit-activeml 0.2.3

We are happy to announce the following changes/features:

  • Bugfix of import error in EpistemicUncertaintySampling due to scikit-learn update
  • Improved compatibility for scikit-learn with version >= 1.1
  • Include unit tests to check compatibility with old dependencies
  • Improved compatibility for Python 3.10
  • Automatic limitations to future package updates in requirements.txt

- Python
Published by mherde almost 4 years ago

scikit-activeml - scikit-activeml 0.2.2

We are happy to announce the following changes/features: - Adding QUIRE to the pool sub package - Restructuring code with black

- Python
Published by mherde almost 4 years ago

scikit-activeml - scikit-activeml 0.2.1

We are happy to announce the following changes/features:

  • EER now supports: substract_current
  • Bug fix: Voi now also working for one unlabeled instance
  • Release Process Improvement: Version number is now automatically set when a new release is drafted

- Python
Published by dakot almost 4 years ago

scikit-activeml - scikit-activeml 0.2.0

We are happy to announce the following changes/features:

  • Speed up documentation build
  • Enhanced presentation on PyPI
  • Optimized PyPI publishing

- Python
Published by mherde almost 4 years ago

scikit-activeml - scikit-activeml 0.1.2

We are happy to announce the following changes/features:

  • Restructure query functions signature in query strategies.
  • Optimized visualization.
  • Code reformatting with black and flake8.
  • Optimization of expected error reduction, including new versions (e.g., VOI).

- Python
Published by mherde almost 4 years ago

scikit-activeml - scikit-activeml 0.1.1

We are happy to announce the following changes/features:

  • Enhanced presentation on PyPI

- Python
Published by dakot over 4 years ago

scikit-activeml - scikit-activeml 0.1.0

We are happy to announce the following changes/features: - Stream-based AL - Multi-annotator AL - Moved clf from __init__ to query for query strategies

- Python
Published by dakot over 4 years ago

scikit-activeml - scikit-activeml 0.0.0

The journey begins...

- Python
Published by dakot almost 5 years ago