Skip to content

Model Training & Adaptation

Feature engineering

Preparing useful clues from information so an AI can learn patterns more easily.

Example

A team derives days since the last purchase from transaction dates.

Why people use it

It helps turn raw records into clues that are more useful for learning.

What you'll hear

“Calculate days since the last purchase from these dates.”

What this means for you

Validate features on data that reflects real prediction conditions.

Can you control it?

Developer-only

The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.

Common questions

Does feature engineering guarantee better performance?
No. New features can be irrelevant, biased or leak future information.
Can a useful-looking clue be unavailable in real use?
Yes. Information recorded only after an event cannot fairly help predict that event beforehand.
Can it make explanations easier?
Sometimes. A meaningful derived measurement may be easier to discuss than many raw records.

Related terms

Still have questions?

Up to 500 characters.

Ask LATHIC about AI. Relevant glossary entries may be included.

Your question, the glossary entries it matches, and a rotating pseudonymous identifier go to Microsoft Azure’s OpenAI service through Vercel AI Gateway to generate an answer. Zero retention and no training are required of the provider, and LATHIC does not save your question or answer. Privacy Notice