v0.24.4

version 0.24.4

• Update LazyDeepMTS: No more LazyMTS class, instead, you can use LazyDeepMTS with n_layers=1
• Specify forecasting horizon in LazyDeepMTS (see updated docs and examples/lazy_mts_horizon.py)
• New class ClassicalMTS for classsical models (for now VAR and VECM adapted from statsmodels) in multivariate time series forecasting (not available in LazyDeepMTS yet)
• partial_fit for CustomClassifier and CustomRegressor

v0.23.1

• Copula simulation in classes MTS and DeepMTS
  • based on copulas of in-sample residuals: vine-tll (default), vine-bb1, vine-bb6, vine-bb7, vine-bb8, vine-clayton, vine-frank, vine-gaussian, vine-gumbel, vine-indep, vine-joe, vine-student
  • sequential split conformal prediction (scp) + vine copula based on calibrated residuals: scp-vine-tll, scp-vine-bb1, scp-vine-bb6, scp-vine-bb7, scp-vine-bb8, scp-vine-clayton, scp-vine-frank, scp-vine-gaussian, scp-vine-gumbel, scp-vine-indep, scp-vine-joe, scp-vine-student
  • sequential split conformal prediction (scp2) + vine copula based on standardized calibrated residuals: scp2-vine-tll, scp2-vine-bb1, scp2-vine-bb6, scp2-vine-bb7, scp2-vine-bb8, scp2-vine-clayton, scp2-vine-frank, scp2-vine-gaussian, scp2-vine-gumbel, scp2-vine-indep, scp2-vine-joe, scp2-vine-student
• cross_val_score: time series cross-validation for MTS and DeepMTS

v0.22.7

• Implement new types of predictive simulation intervals (parameter type_pi) in class MTS: independent bootstrap, block bootstrap, 2 variants of split conformal prediction:
  - gaussian: simple, fast, but: assumes stationarity of Gaussian in-sample residuals and independence in the multivariate case
  - kde: based on Kernel Density Estimation of in-sample residuals
  - bootstrap: based on independent bootstrap of in-sample residuals
  - block-bootstrap: based on basic block bootstrap of in-sample residuals
  - scp-kde: Split conformal prediction with Kernel Density Estimation of calibrated residuals
  - scp-bootstrap: Split conformal prediction with independent bootstrap of calibrated residuals
  - scp-block-bootstrap: Split conformal prediction with basic block bootstrap of calibrated residuals
  - scp2-kde: Split conformal prediction with Kernel Density Estimation of standardized calibrated residuals
  - scp2-bootstrap: Split conformal prediction with independent bootstrap of standardized calibrated residuals
  - scp2-block-bootstrap: Split conformal prediction with basic block bootstrap of standardized calibrated residuals

• Implement Winkler score in LazyMTS and LazyDeepMTS for probabilistic forecasts

• Use conformalized Estimators in MTS (see examples/mts_conformal_not_sims.py)

• Include block_size for block bootstrapping methods for *MTS classes

v0.20.0

v0.20.0

Prediction intervals using conformal simulation

• See this example, and this notebook

v0.18.1

Prediction intervals using Bayesian inference and conformal prediction

• Bayesian CustomRegressor (for sklearn's BayesianRidge, ARDRegressor, and GaussianProcessRegressor)

• Conformalized CustomRegressor (splitconformal and localconformal for now): see examples and notebook

• See this example, this example, and this notebook

v0.17.2

• Attribute estimators (a list of Estimator's as strings) for LazyClassifier,
  LazyRegressor, LazyDeepClassifier, LazyDeepRegressor, LazyMTS, and LazyDeepMTS
• New documentation for the package, using pdoc (not pdoc3)
• Remove external regressors xreg at inference time for MTS and DeepMTS
• New class Downloader: querying the R universe API for datasets (see
  https://thierrymoudiki.github.io/blog/2023/12/25/python/r/misc/mlsauce/runiverse-api2 for similar example in mlsauce)
• Add custom metric to Lazy*
• Rename Deep regressors and classifiers to Deep* in Lazy*
• Add attribute sort_by to Lazy* -- sort the data frame output by a given metric
• Add attribute classes_ to classifiers (ensure consistency with sklearn)
• Add preprocessing to all LazyDeep*

v0.16.5

• Improve consistency with sklearn's v1.2, for OneHotEncoder
• add robust scaler for creating layers and clustering
• relatively faster scaling in preprocessing
• Regression-based classifiers (see https://www.researchgate.net/publication/377227280_Regression-based_machine_learning_classifiers)
• DeepMTS (multivariate time series forecasting with deep quasi-random layers): see https://thierrymoudiki.github.io/blog/2024/01/15/python/quasirandomizednn/forecasting/DeepMTS
• AutoML for MTS (multivariate time series forecasting): see https://thierrymoudiki.github.io/blog/2023/10/29/python/quasirandomizednn/MTS-LazyPredict
• AutoML for DeepMTS (multivariate time series forecasting): see https://github.com/Techtonique/nnetsauce/blob/master/nnetsauce/demo/thierrymoudiki_20240106_LazyDeepMTS.ipynb
• Subsample continuous and discrete responses

v0.16.3

• add robust scaler
• relatively faster scaling in preprocessing
• Regression-based classifiers (see https://www.researchgate.net/publication/377227280_Regression-based_machine_learning_classifiers)
• DeepMTS (multivariate time series forecasting with deep quasi-random layers): see https://thierrymoudiki.github.io/blog/2024/01/15/python/quasirandomizednn/forecasting/DeepMTS
• AutoML for MTS (multivariate time series forecasting): see https://thierrymoudiki.github.io/blog/2023/10/29/python/quasirandomizednn/MTS-LazyPredict
• AutoML for DeepMTS (multivariate time series forecasting): see https://github.com/Techtonique/nnetsauce/blob/master/nnetsauce/demo/thierrymoudiki_20240106_LazyDeepMTS.ipynb
• Subsample continuous and discrete responses

v0.16.2

• AutoML for MTS (multivariate time series forecasting)
• AutoML for DeepMTS (multivariate time series forecasting)
• Subsample continuous and discrete responses
• Regression-based classifiers

v0.15.0

Automated Machine learning (AutoML)

• lazy predict for classification and regression (see https://thierrymoudiki.github.io/blog/2023/10/22/python/quasirandomizednn/nnetsauce-lazy-predict-preview)
• lazy predict for multivariate time series (see https://thierrymoudiki.github.io/blog/2023/10/29/python/quasirandomizednn/MTS-LazyPredict)
• lazy predict for deep quasi-randomized classifiers and regressors nnetworks (see this example for classification: examples/lazy_custom_deep_classification.py; and this example for regression: examples/lazy_custom_deep_regression.py