Issues resolved

• Update minimum test images – Issue #563 by @sarahmish
• Update benchmark – Issue #538 by @sarahmish
• Pin MacOS Version in GH Actions – Issue #535 by @sarahmish
• Add anomaly transformer method – Issue #452 by @sarahmish

Support for python 3.10 and 3.11

Issues resolved

• Update test_core file – #507 by @sarahmish
• update ARIMA primitive and pipeline – #503 by @sarahmish
• Update Dependency – #497 & #499 by @sarahmish
• Update Python for Dependency Test – #484 by @sarahmish
• Add python 3.10 and 3.11 & drop 3.6 and 3.7 – #477 by @sarahmish
• LNN Pipeline – #475 by @sarahmish

Support for python 3.9 and new Matrix Profile pipeline

Issues resolved

• Send pipeline names in benchmark arguments – Issue #466 by @sarahmish
• Add continuation parameter for benchmark – Issue #464 by @sarahmish
• Fix references in documentation – Issue #453 by @sarahmish
• Update documentation – Issue #448 by @sarahmish
• Add matrix profiling method – Issue #446 by @sarahmish
• Support python 3.9 – Issue #408 by @sarahmish

This version introduces a new dataset to the benchmark.

Issues resolved

• Add UCR dataset to the benchmark – Issue #443 by @sarahmish
• docker image build failed – Issue #439 by @sarahmish
• Edit interval settings in azure pipeline – Issue #436 by @sarahmish

This version uses ml-stars package instead of mlprimitives.

Issues resolved

• Migrate to ml-stars – Issue #418 by @sarahmish
• Updating best_cost in find_anomalies primitive – Issue #403 by @sarahmish
• Retire lstm_dynamic_threshold_gpu and lstm_autoencoder_gpu pipeline maintenance – Issue #373 by @sarahmish
• Typo in xlsxwriter dependency specification – Issue #394 by @sarahmish
• orion.evaluate uses fails when fitting – Issue #384 by @sarahmish
• AER pipeline with visualization option – Issue #379 by @sarahmish

Issues resolved

• Move VAE from sandbox to verified – Issue #377 by @sarahmish
• Pin opencv – Issue #372 by @sarahmish
• Pin scikit-learn – Issue #367 by @sarahmish
• Fix VAE documentation – Issue #360 by @sarahmish

This version introduces several new enhancements:

• Support to python 3.8
• Migrating to Tensorflow 2.0
• New pipeline, namely VAE, a Variational AutoEncoder model.

Issues resolved

• Add python 3.8 – Issue #342 by @sarahmish
• VAE (Variational Autoencoders) pipeline implementation – Issue #349 by @dyuliu
• Add masking option for regression_errors – Issue #352 by @dyuliu
• Changes in TadGAN for tensorflow 2.0 – Issue #161 by @lcwong0928
• Add an automatic dependency checker – Issue #320 by @sarahmish
• TadGAN batch_size cannot be changed – Issue #313 by @sarahmish

This version fixes some of the issues in aer, ae, and tadgan pipelines.

Issues resolved

• Fix AER model predict error after loading – Issue #304 by @lcwong0928
• Update AE to work with any window_size – Issue #300 by @sarahmish
• Updated tadgan_viz.json – Issue #292 by @Hramir

This version introduce a new pipeline, namely AER, an AutoEncoder Regressor model.

Issues resolved

• Add AER Model - Issue #286 by @lcwong0928

This version deprecates the support of OrionDBExplorer, which has been migrated to
sintel. As a result, Orion no longer requires
mongoDB as a dependency.

Issues resolved

• Update dependency - Issue #283 by @sarahmish
• General housekeeping - Issue #278 by @sarahmish
• Fix tutorial testing issue - Issue #276 by @sarahmish
• Migrate OrionExplorer to Sintel - Issue #275 by @dyuliu
• LSTM viz JSON pipeline added - Issue #271 by @Hramir