iterative broadening

Terms from Artificial Intelligence: humans at the heart of algorithms

Iterative broadening is a variant of breadth-first search designed to deal with search trees where some nodes have very large or even infinite numbers of children (large branching factor). It proceeds just like normal breadth-first search, but limits its search to the first B children of a node, where B is a limiting factors, perhaps 5 children. If this search is unsuccessful, B is increased and the process repeated (with tweaks to avoid repeated work).

Defined on page 68

Used on pages 68, 80