Skip to content

Releases: DataCanvasIO/HyperTS

Release 0.2.2

22 Dec 02:55
Compare
Choose a tag to compare

Details of the HypertTS update are as follows:

  • Adapted to TensorFlow versions 2.11.0 and above.

  • Fixed the issue of unstable random seed in NAS mode.

  • Fixed handling of int variables in DL mode.

  • Updated CI.

  • Fixed the unordered indexing problem in the arrow_head data loading.

  • Refactored the identification of discrete variables in DL mode.

  • Added the Lion optimizer.

  • Corrected variable name spelling error, HybirdRNN -> HybridRNN.

  • Corrected the trend parameter in the VAR model.

  • Adjusted the field of view length for evaluation segmentation in forecast tasks.

  • Added seasonal analysis.

  • Fixed the issue of truncating negative values to 0 in forecast tasks.

  • Updated hypernets and numpy versions.

  • Supported the AdamW optimizer (tensorflow >= 2.14.0).

  • Adjusted the legend margin in visualizations.

  • Add TimeSeriesForestClassifier and IndividualTDEClassifier built-in algorithm for stats mode classification.

  • sktime dependencies are removed and replaced with built-in algorithm support. If need to use KNN classification, please manually install sktime.

  • Thanks to @NatLee for his contributions to hyperts.

Release 0.2.0

24 Nov 09:43
Compare
Choose a tag to compare

Details of the HypertTS update are as follows:

  • Supported time series anomaly detection task, and adapt to the full pipeline automation process;

  • Added IForest anomaly detection model (stats mode);

  • Added TSOneClassSVM anomaly detection model(stats mode);

  • Added ConvVAE anomaly detection model(dl mode);

  • Added realKnownCause anomaly detection dataset;

  • Supported the visualization of anomaly detection results, and can analyze the anomaly location and severity;

  • Compatible with Prophet version 1.1.1, now pip install hyperts for simultaneous successful prophet installation;

  • Compatible with all versions of scipy;

  • Added API documentation module;

  • Supported for model persistence (saving and reloading trained models);

  • In ```model.predict()``, fixed missing value handling;

  • For the time series forecast task, the forecast function of DL model is calibrated;

  • DLClassRegressSearchSpace was refactored for better adaptation to regression task;

  • Extend InceptionTime to solve the regression task;

  • Fixed some known issues;

  • Thanks to @peter Cotton for his contributions to hyperts.

Release 0.1.4

22 Jul 07:40
Compare
Choose a tag to compare

See 0.1.3.

Release 0.1.3

19 Jul 04:12
Compare
Choose a tag to compare

Details of the HypertTS(version 0.1.3) update are as follows:

  • Tuning search space hyperparameters;
  • Added report_best_trial_params;
  • Fixed ARIMA to be compatible with version 0.12.1 and above;
  • Fixed the pt issue of LSTNet;
  • Simplified custom search space, task, timestamp, covariables and metircs can not be passed;
  • Added OutliersTransformer, supported dynamic handling of outliers;
  • Adjusted final train processing - lr, batch_size, epcochs and so on;
  • Added time series meta-feature extractor;
  • Added Time2Vec, RevIN, etc. layers;
  • Added N-Beats time series forecasting model;
  • Added InceptionTime time series classification model;
  • Supported dynamic downsampling for time series forecasting;
  • Refactored positive label inference method;
  • Added neural architecture search mode;
  • Fixed some known issues

Release 0.1.2.1

25 Apr 07:45
Compare
Choose a tag to compare

1、Supports cross validation.
2、Supports greedy ensemble.
3、Supports time series forecasting data without timestamp column.
4、Supports time series forecasting truncation training set to train.
5、Supports time series forecasting of discrete data (no time frequency).
6、Supports Fourier inference period.
7、Optimizes search space and architecture.
8、Fixes some bugs.

0.1.1

08 Apr 03:06
Compare
Choose a tag to compare
Release 0.1.1