Evaluating Time Series Models
Evaluating time series models is usually different from most other machine learning tasks as we usually don’t have i.i.d. data.
Out-of-sample
![Out-of-Sample with Sliding Window](../assets/evaluate-time-series-models/oos-sliding-window.png)
Out-of-Sample with Sliding Window
If the sliding window size is 1, then we have the simplest out-of-sample holdout scenario.
Prequential
![Prequential with Gap](../assets/evaluate-time-series-models/block-gap.png)
Prequential with Gap
![Prequential with Growing Train](../assets/evaluate-time-series-models/block-grow.png)
Prequential with Growing Train
![Prequential with Sliding Blocks](../assets/evaluate-time-series-models/block-slide.png)
Prequential with Sliding Blocks
Cross-validation
![Cross-validation](../assets/evaluate-time-series-models/cv.png)
Cross-validation
![Cross-validation with Neighbor removed](../assets/evaluate-time-series-models/cv-removed.png)
Cross-validation with Neighbor removed
Planted:
by L Ma;
Similar Articles:
L Ma (2022). 'Evaluating Time Series Models', Datumorphism, 04 April. Available at: https://datumorphism.leima.is/wiki/forecasting/evalutate-time-series-models/.