feature-engineering

What problem does it solve?

This Skill automates the creation of informative features for tabular datasets, enabling faster model development and better predictive performance by systematically applying a wide range of feature engineering techniques.

Core Features & Use Cases

• Target encoding and mean encoding for high-cardinality categories
• Interaction features: products, ratios, and differences between key variables
• Time-based features: day, month, year, day_of_week, hour, quarter
• Aggregation features: mean, std, min, max, median, sum, count
• Automated feature synthesis using tools like Deep Feature Synthesis (DFSynth)
• Handling missing values with indicators and imputation-aware features
• Text and numerical feature vectorization for richer representations

Quick Start

Use the feature-engineering skill to generate target-encoded features, interaction terms, and time-based aggregations for the dataset 'sales.csv'.