Index

Definition

Feature exposure (more specifically, max feature exposure) is a measure of how well balanced a model’s exposure is to the features. Models with lower feature exposures tend to have more consistent performance over the long run.

<aside> 💡 Feature exposure and max feature exposure are similar but not exactly the same

</aside>

Code Snippet

You can add code notation to any Notion page:

• Type /code and press enter.
• Choose the language from the dropdown in the bottom right corner.
• Here's an example:

def feature_exposures(df):
    feature_names = [f for f in df.columns
                     if f.startswith("feature")]
    exposures = []
    for f in feature_names:
        fe = spearmanr(df[PREDICTION_NAME], df[f])[0]
        exposures.append(fe)
    return np.array(exposures)

Citations & Other Resources

Numerai Forums

Model Diagnostics: Feature Exposure

An introduction to feature neutralization / exposure

Other Resources