Model Training & Adaptation
Feature selection
Choosing which pieces of information an AI will use to make predictions.
Example
A team removes irrelevant fields before training a prediction system.
Why people use it
It removes information that adds little value or makes learning harder.
What you'll hear
“Do we need every field in this spreadsheet?”
What this means for you
Check that selection used training data rather than information from the test set.
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
- Should clue selection use the final test answers?
- No. That would give the design an unfair preview and make the final check less independent.
- Does more information always improve the AI?
- No. Irrelevant or misleading details can make it harder to learn useful patterns.
- Can removing a field improve privacy?
- It can reduce what is collected or used, though other fields may still reveal sensitive information.