Feature engineering involves choosing which attributes of a return to feed into the model. For text-based classification, features might include word frequency, text length, presence of specific keywords, or sentiment scores. For structured data, features could be product category, price point, or customer tenure. Good feature engineering significantly improves model performance and reduces the data needed for training.
← Glossary
Feature Engineering
The process of selecting and transforming raw data into the most useful inputs for an AI model.
Related terms
- Training Data — The labeled examples used to teach an AI model how to categorize returns correctly.
- Natural Language Processing — A branch of AI that helps computers understand, interpret, and generate human language.
- Machine Learning Model — A mathematical system trained on historical data to make predictions or classifications on new data.
- Labeled Data — Training examples where humans have already assigned the correct category or answer.