Tuning Parameters and Scoring Backtests in BigQuant
Summary
The post shows a BigQuant parameter-tuning setup in which a grid of candidate values is supplied for a model parameter, with examples of feature and tree-count settings. It then defines a scoring function that reads the final Sharpe value from a backtest result and returns it as the score used to compare parameter choices.
The exchange does not establish that this setup is correct or explain why the reported output seems wrong. The response asks whether the referenced backtest module produces a result and requests the code for diagnosis. As a result, the post offers a basic illustration of tuning inputs and a Sharpe-based objective, but gives no resolved troubleshooting steps, performance evidence, or guidance on validation and overfitting.
Key ideas
- A parameter grid can list candidate values for model or feature settings.
- A scoring function can extract the final backtest Sharpe value for optimization.
- The post leaves the reported tuning issue unresolved and asks for the backtest code to investigate.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.