Premium
MODELING MODELING
Author(s) -
Killeen Peter R.
Publication year - 1999
Publication title -
journal of the experimental analysis of behavior
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.75
H-Index - 61
eISSN - 1938-3711
pISSN - 0022-5002
DOI - 10.1901/jeab.1999.71-275
Subject(s) - consistency (knowledge bases) , computer science , goodness of fit , set (abstract data type) , function (biology) , task (project management) , range (aeronautics) , feature (linguistics) , cascade , artificial intelligence , machine learning , engineering , linguistics , philosophy , systems engineering , evolutionary biology , chemical engineering , biology , programming language , aerospace engineering
Models are tools; they need to fit both the hand and the task. Presence or absence of a feature such as a pacemaker or a cascade is not in itself good. Or bad. Criteria for model evaluation involve benefit‐cost ratios, with the numerator a function of the range of phenomena explained, goodness of fit, consistency with other nearby models, and intangibles such as beauty. The denominator is a function of complexity, the number of phenomena that must be ignored, and the effort necessary to incorporate the model into one's parlance. Neither part of the ratio can yet be evaluated for MTS, whose authors provide some cogent challenges to SET.