Reusing Raw Features for Rolling-Window Model Processing

Summary

This workflow note explains how to prepare data when training models across rolling windows. As each window advances, the training sample changes, and learned processor state—such as means and standard deviations—must be recalculated for that window. Reusing processed features from another window could therefore leave the model with stale transformations.

The proposed approach loads raw features that do not depend on the rolling window through a DataHandler-based DataLoader, then applies processors to generate features whose transformations depend on the current window. This is intended to avoid regenerating the window-independent data while keeping rolling-dependent processing current. The document provides only a brief workflow description and a command for running an example; it supplies no code details, timing comparison, or model performance evidence.

Key ideas

• Rolling training windows change both the training data and processor state.
• Window-dependent transformations such as means and standard deviations need to be refreshed as the window moves.
• The workflow loads reusable raw features with a DataHandler-based DataLoader.
• Processors then generate the features that depend on the current rolling window.
• The note describes an example but reports no efficiency or model-performance results.

Tags

# Rolling Process Data


# Rolling Process Data

This workflow is an example for `Rolling Process Data`.

## Background

When rolling train the models, data also needs to be generated in the different rolling windows. When the rolling window moves, the training data will change, and the processor's learnable state (such as standard deviation, mean, etc.) will also change. 

In order to avoid regenerating data, this example uses the `DataHandler-based DataLoader` to load the raw features that are not related to the rolling window, and then used Processors to generate processed-features related to the rolling window.


## Run the Code

Run the example by running the following command:
```bash
    python workflow.py rolling_process
```

Shown in full with attribution under the source's licence. Licence: MIT

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.