uncertainty measure

Terms from Artificial Intelligence: humans at the heart of algorithms

The glossary is being gradually proof checked, but currently has many typos and misspellings.

Given uncertainty of sensors and uncertan reasoning we often want to have some idea of the level of certainty of the results we are seeing. Examples in statistics include the standard deviation or confidence intervals, and there are parallels in AI algorithms. This may be used in further processing (for example giving additional weight to more certain values) or for visualising uncertainty (for example using error bars on graphs). Of course any such measure of uncertainty is itself usually uncertain!

Used in Chap. 7: page 88

Also known as measure of uncertainty