2011年5月1日
Naive and yet Enlightened: From Natural Frequencies to Fast and Frugal Decision Trees
Heuristics: The Foundations of Adaptive Behavior
- ,
- ,
- ,
- 記述言語
- 英語
- 掲載種別
- 論文集(書籍)内論文
- DOI
- 10.1093/acprof:oso/9780199744282.003.0006
- 出版者・発行元
- Oxford University Press
The trees for classification and for decision that are introduced in this chapter are naïve, fast, and frugal. Why they are "naïve," that is, why they ignore conditional dependencies between cues, is extensively illustrated. Why they are "frugal," in the sense that they tend to use much fewer cues than those provided by the environment, is also explained in detail. The trees' "fastness" appears as a consequence of their frugality. The main property of these trees is that they implement one-reason classification/decision in analogy with heuristics for one-reason decision making. The lexicographic nature of these trees is shown with the help of a characterization theorem. As will also become clear, fastand-frugal trees generalize take-the-best.
- ID情報
-
- DOI : 10.1093/acprof:oso/9780199744282.003.0006
- SCOPUS ID : 84865200151