Experimental

API details

Change-point utilities:

change_points.get_ruptures_regularization(ts, ...)

Get regularization parameter values for given number of changepoints.

Classification of time-series:

classification.TimeSeriesBinaryClassifier(...)	Class for holding time series binary classification.
classification.PredictabilityAnalyzer(...[, ...])	Class for holding time series predictability prediction.
classification.feature_extraction.TSFreshFeatureExtractor([...])	Class to hold tsfresh features extraction from tsfresh.
classification.feature_extraction.WEASELFeatureExtractor(...)	Class to extract features with WEASEL algorithm.

Prediction Intervals:

prediction_intervals.BasePredictionIntervals(...)	Base class for prediction intervals methods.
prediction_intervals.NaiveVariancePredictionIntervals(...)	Estimate prediction variance based on historical residuals.
prediction_intervals.ConformalPredictionIntervals(...)	Estimate conformal prediction intervals using absolute values of historical residuals.
prediction_intervals.EmpiricalPredictionIntervals(...)	Estimate prediction intervals using values of historical residuals.

Prediction Intervals utilities:

prediction_intervals.utils.residuals_matrices(...)

Estimate residuals matrices with backtest.