v0.8.2

v0.8.2 (2024-12-04)

Fix

• fix: fix release.yml (6be2832)

v0.8.1

v0.8.1 (2024-12-04)

Fix

• fix: update docker release (af93fca)

v0.8.0

v0.8.0 (2024-12-01)

Build

• build: use python 3.11, 3.12, and 3.13 (993b12a)

Chore

• chore: updating moa jar 24-09-08 (b7e958c)

Ci

• ci: fix docker triggers (2759800)

• ci: fix automated docker versioning (0cc493e)

Documentation

• docs: add autoML notebook with examples (#198) (ac7e578)

• docs: update csmote init doc (0963190)

• docs: fix broken link in README (40e98d0)

Feature

• feat: add optional progress bars (0610c4b)

• feat: abcd, a drift detector for univariate and multivariate data (5c75c24)

• feat: add shrub ensembles for online classification

Buschjäger, S., Hess, S., & Morik, K. J. (2022, June). Shrub ensembles for
online classification. In Proceedings of the AAAI Conference on Artificial
Intelligence (Vol. 36, No. 6, pp. 6123-6131). (f966e32)

• feat: added wrapper code for 6 stream generators (ee2cd30)

• feat: add autoclass for automl (b5adc95)

• feat: added code for led and agrawal stream generators (3356f00)

• feat: added drift object capabilities to arf (f792667)

• feat: added weightedknn (e9c5757)

• feat: create synthetic data stream for anomaly detection

Co-authored-by: Heitor Murilo Gomes <[email protected]> (d1dbd78)

• feat: clustering, 4 wrappers, notebook and documentation (b5c8f28)

• feat: clustering methods, documentation and notebook (dc68be6)

Fix

• fix: fix numpy 2.0.0 compatibility (#199) (7bbaa96)

• fix: ensuring deterministic behavior for driftstream (33ba7ba)

• fix: fixing documentation for notebooks toctree (21fb4aa)

• fix: adding gifs shown in notebook (6c08d39)

Test

• test: stop skipping notebooks in CI/CD pipeline (dfcd0c6)

Unknown

• Pipelines V2 (#200)

Co-authored-by: Heitor Murilo Gomes <[email protected]> (09a80eb)

v0.7.0

v0.7.0 (2024-08-03)

Ci

• ci: fix docker versioning (bbfc8b5)

Documentation

• docs: update plotting to prediction interval notebook (f33c3d1)

Feature

• feat: clustering base classes with simple visualization

Co-authored-by: Heitor Murilo Gomes <[email protected]> (09b3a60)

Fix

• fix: update visualization for compatibility (e25f6b6)

v0.6.0

v0.6.0 (2024-07-31)

Ci

• ci: fix workflow trigger of docker build (d8fdd55)

Feature

• feat: add restart_stream=False as an option in evaluators (a965dc4)

• feat: add clustering and ssl (ae8b592)

• feat: add IJCAI 2024 tutorial (cac7900)

Fix

• fix: remove tutorial notebook (d82e05d)

• fix: add overwritten changes in commit cc13390 for recurrent concept drifts (472a597)

v0.5.0

v0.5.0 (2024-07-30)

Ci

• ci: fix bad condition (b096339)

• ci: trigger docker builds on release (1ef2ad6)

Documentation

• docs: recommend https git clone (5c9715a)

• docs: fix __init__ documentation (29c516a)

• docs(about.rst): add Justin Liu (4d97ead)

Feature

• feat: add reference to Autoencoder anomaly detector (d8787e8)

• feat(Anomaly-Detection): add Autoencoder anomaly detector (e27cbe9)

Fix

• fix: updating AD docs and notebook (47ee056)

• fix: corrections to anomaly evaluation (57b93c3)

v0.4.0

v0.4.0 (2024-07-25)

Build

• build: ignore the stub build issue in editable installs (347ed74)

Documentation

• docs(tutorial.rst): add talks section (c5a339b)

• docs(drift): update drift detector docs (4c69478)

• docs: rebase (748cc94)

• docs: use sphinx.ext.autosummary to generate more complete api docs (01d270f)

• docs: add about us page (f3af4d5)

Feature

• feat(Anomaly-Detection): add cumulative and prequential evaluation for anomaly (a8124bc)

• feat: updated evaluation (cc13390)

• feat: add recurrent concept drift API (24327fc)

• feat: drift detection API (50d76ad)

• feat(anomaly): add OnlineIsolationForest anomaly detector (5a57bb0)

• feat: add anomaly_threshold and size_limit parameters to HalfSpaceTree (21804d0)

• feat(Misc): add unit tests for model save and load functionality (739f8af)

Fix

• fix: updating evaluation to use prequential results (2b6abab)

• fix: fix issue saving and loading models (04d9207)

Test

• test: speed up parallel_ensembles.ipynb (e9a6bb0)

• test: rename name of unit test (4dbecb3)

v0.3.1

v0.3.1 (2024-06-10)

Fix

• fix: updating tutorials formatting (36a0a29)

v0.3.0

v0.3.0 (2024-05-30)

Build

• build: add gcc and g++ to docker for arm support (3ede9cd)

• build: build.clean-moa now includes build.clean-stubs (afcc080)

• build: fix dockerfile base (88b7b19)

• build: add docker arm64 (7f1d25f)

• build: add docker (e07288c)

Documentation

• docs: adding documentation for regressors (64c7105)

• docs(Anomaly-Detection): add anomaly detection in the extra tutorials section (3bc6f94)

• docs: fix benchmarking link (06202f8)

• docs: update installation.md (9324fd4)

Feature

• feat: parallel ensembles (#132)

Co-authored-by: Heitor Murilo Gomes <[email protected]> (1a3e6a3)

• feat(Misc): export and import models (a1af686)

• feat: adding LeveragingBagging and OnlineAdwinBagging classes (25f0f75)

• feat(pipelines): updated moa jar and minor changes (e71b065)

• feat(pipelines): updated notebook introducing the concepts (4c787d6)

• feat(Anomaly-Detection): add abstract class for anomaly detector algorithms (df74b25)

• feat(STREAM): change the logic when dealing with csv input (2c2a89d)

• feat(CSMOTE): adding CSMOTE for imbalanced learning (e39d006)

• feat(Anomaly-Detection): add anomaly detection notebook (f0b6d4f)

• feat(Anomaly-Detection): add Half-Space Trees

Add wrapper for Half-Space Trees for MOA. (aa85c0f)

• feat: evaluation tests and fixes (147612b)

• feat: update plotting regression/prediction interval (bf3f7ce)

• feat: add HAT, DWM, and SAMkNN wrappers (314db2e)

• feat: add plot_regression/prediction_interval_results functions (fe44e0f)

• feat: add wrapper for hyperplane (c and r) generators (6cd23af)

• feat: add prediction interval notebook (f64a2e1)

Fix

• fix: updating save_load notebook (c4e11b0)

• fix(Misc): add a title for tutorial of save and load model (eaad513)

• fix: adding tests for new ensembles (7ecd80d)

• fix: expanding example notebook (105bf4c)

• fix: add notebook 07 to toctree (d527e29)

• fix: moajar hash (e02d8ac)

• fix: add pipelines notebook to invoke.yml (800ed73)

• fix: call prepareForUse() in initializer of stream when schema is provided (9f29e61)

• fix: corrected docstring for stream_from_file (e627f2d)

• fix: updated doc and ARFF exception handling (187c429)

• fix: fix stream consistency (f363788)

• fix: force target as int when categorical (bc61941)

• fix: update from_custom function in _stream.py (1bbba9b)

• fix: update PytorchStream.py with target_type (d8e9446)

• fix: update instance.py for target_type (e602dc3)

• fix: update test_batch to enforce classification in basic test (bda30f5)

• fix: small alteration (d8c44d8)

• fix: remove None| and update test_regressor (2cc3ba5)

• fix: refine documentation for SOKNl and SOKNLBT (1c013f0)

• fix: fix the end-strings (bf114cf)

• fix: update evaluation.py (bf62b0e)

• fix: update evalutation.py (6e1e90f)

• fix: update moa-url and sha256 value (491f617)

• fix: allow _prequential_evaluation_fast to store y and predictions (9675615)

• fix: correcting author name in ref (991f916)

• fix(Anomaly-Detection): change expect output of example and test code for Half-Space Trees (5c0832a)

• fix: updating HalfSpaceTrees example usage (fa1e236)

• fix: updating moa jar 24-05-20 (3327f6b)

• fix(Anomaly-Detection): change expect output for Half-Space Trees for unit testing (6b607c3)

• fix(Anomaly-Detection): change expect output for the code example of Half-Space Trees (2f7225a)

• fix: test 00 notebook (771d784)

• fix: updating ignored notebooks (47e174a)

• fix: remove useless code in PI evaluator (ea500af)

• fix: fine-tune transparency for prediction interval plotting (62c19a5)

• fix: add set_context in base and remove it for samknn ([4baad0f](https://github.com/adaptive-machine-learning/...

v0.2.1

v0.2.1 (2024-05-06)

Documentation

• docs: add blurb to plot and improve moajar update steps (cd6ca55)

• docs: use wall time in benchmark plot (31bb320)

Fix

• fix: tweak docs and add links between sites (5eca95d)