Prompt

Design feature engineering pipeline for ML model.

<raw_data>
{describe available raw features}
</raw_data>

<target_variable>
{what you're predicting}
</target_variable>

<domain_knowledge>
{business context and rules}
</domain_knowledge>

<model_requirements>
{interpretability, performance needs}
</model_requirements>

Engineer features:
1. Exploratory analysis
   - Feature distributions
   - Target correlations
   - Multicollinearity check
   
2. Transformation strategies
   - Numerical scaling/normalization
   - Categorical encoding options
   - DateTime extractions
   - Text vectorization
   
3. Feature creation
   - Domain-specific features
   - Interaction terms
   - Polynomial features
   - Aggregations
   
4. Feature selection
   - Filter methods
   - Wrapper methods
   - Embedded methods
   - Dimensionality targets
   
Include feature importance analysis.