Data science
Feature engineering
Transform raw data into model-ready features with encoding, scaling, and automated creation.
Feature engineering turns raw data into inputs that improve model performance. Sourcetable automates common transformations through the AI assistant.
Creates: year, quarter, month, week, day of week, day of month, hour, minute, is_weekend, is_holiday, days_since_start.
Group-by aggregations: sum, mean, median, count, min, max, std, first, last.
Products, ratios, and polynomial features between numeric columns.
Creates: word count, character count, average word length, sentiment score, TF-IDF vectors.
Trains a model and ranks features by their contribution to predictions.
Calculates VIF for each feature. Values above 5-10 indicate problematic multicollinearity.
Iteratively removes the least important feature until performance stops improving.