incremental induction

incremental induction

[‚iŋ·krə′ment·əl in′dək·shən]
(electromagnetism)
The quantity lying between the highest and lowest value of a magnetic induction at a point in a polarized material, when subjected to a small cycle of magnetization.
References in periodicals archive ?
5 [Quinlan 1993]; AQ, CN2 [Clark and Niblett 1989]; AE1 [Hong 1985], and HCV [Wu 1993d]), incremental induction algorithms (such as ID5R [Utgoff 1989], and the version-space method [Mitchell 1977]) fall into the supervised classification category.
Incremental induction in the case of large, dynamic real-world databases: There are several common problems in all kinds of inductive learning algorithm: (1) When a database is very large, how can these algorithms speed up their learning processes?
Generally speaking, incremental induction can take more time (but less run-time space) because it needs to restructure decision trees or rules when some new examples do not fit the decision trees or rules developed so far.
Full browser ?