Where to Apply Standardization in BigQuant’s DNN Template

Summary

This Chinese-language support exchange asks why BigQuant’s DNN template does not show a separate standardization module and how to handle an error after adding one. The response explains that the DNN is implemented as code and that data processing, including standardization, is handled by DAI by default. It advises placing standardization code in the input-features DAI SQL component.

The exchange is platform-specific guidance about where preprocessing belongs in this workflow. It does not explain a particular scaling method, how to fit preprocessing without leaking information across training and validation periods, or how to diagnose the reported error. No model results or trading evidence are included, so the note is useful for locating the preprocessing step rather than evaluating a strategy or DNN approach.

Key ideas

• The DNN template is implemented as code rather than as a separate visual standardization module.
• The response says data processing is handled by DAI by default.
• Custom standardization code should be placed in the input-features DAI SQL component.
• The exchange does not specify a scaling formula or explain how to resolve the reported error.
• It provides platform workflow guidance, not evidence about model or trading performance.

Tags

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