threshold

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

A threshold is a critical value beyond which some event occurs or values change drastically. In biological neorones and artificial neurons there is often no reaction for small or moderate amounts of activation from other neurons. However, if the level of input activation goes beyond some threshold, then the neuron will fully fire.

Used in Chap. 6: pages 84, 86, 87, 88, 90; Chap. 9: pages 130, 131, 137, 138; Chap. 10: pages 146, 147; Chap. 12: pages 175, 176, 177, 180, 184, 185, 191, 196, 202; Chap. 19: page 329; Chap. 20: page 335; Chap. 21: page 357

Also known as thresholding