Glossary of Terms for Explainable AI
Clever Hans
Human in the Loop
Level of Risk
In our course, we define five levels of risk related to making a prediction and needing a clear explanation of why a prediction was made.
- Passive Modification: AI agents perform a background process that modifies resources but does not involve high-stakes
- Assisted Decision-Making AI agents provide low-stakes recommendations where human confirmation is required before any action is taken.
- Opt-In High-Stakes: A high-stakes recommendation and humans must opt-in to accept a recommendation made by an AI agent
- Opt-Out High-Stakes: A high-stakes recommendation humans must opt-out to avoid a recommendation
- Autonomous High-Stakes: A high-stakes recommendation and there are no humans in the loop
LIME
Locked Algorithm
An algorithm that provides the same result each time the same input is applied to it and does not change with use. Examples of locked algorithms are static look-up tables, decision trees, and complex classifiers.