Out-of-sample methods

timecave.validation_methods.OOS

This module contains all the Out-of-Sample (OOS) validation methods supported by this package.

Classes:

Name	Description
Holdout	Implements the classic Holdout method.
RepeatedHoldout	Implements the Repeated Holdout approach.
RollingOriginUpdate	Implements the Rolling Origin Update method.
RollingOriginRecalibration	Implements the Rolling Origin Recalibration method.
FixedSizeRollingWindow	Implements the Fixed-size Rolling Window method.

See also

Prequential methods: Prequential or forward validation methods for time series data.

Cross-validation methods: Cross-validation methods for time series data.

Markov methods: Markov cross-validation method for time series data.

Notes

Out-of-sample methods are one of the three main classes of validation methods for time series data (the others being prequential methods and cross-validation methods). Unlike cross-validation methods, this class of methods preserves the temporal order of observations, although it differs from prequential methods in that it does not partition the series into equally sized folds. For a comprehensive review of this class of methods, the reader should refer to [1].

References

1

Leonard J Tashman. Out-of-sample tests of forecasting accuracy: an analysis and review. International journal of forecasting, 16(4):437–450, 2000.