Encyclopedia of Machine Learning

2010 Edition
| Editors: Claude Sammut, Geoffrey I. Webb


  • Claude Sammut
Reference work entry
DOI: https://doi.org/10.1007/978-0-387-30164-8_327
A hypothesis, h, is a predicate that maps an instance to true or false. That is, if h( x) is true, then x is hypothesized to belong to the concept being learned, the target. Hypothesis, h 1, is more general than or equal to h 2, if h 1 covers at least as many examples as h 2 (Mitchell, 1997). That is, h 1h 2 if and only if
$$ (\forall x)[{h}_{1}(x) \rightarrow {h}_{2}(x)]$$
This is a preview of subscription content, log in to check access.


  1. Mitchell, T. M. (1997). Machine learning. New York: McGraw-Hill.MATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2011

Authors and Affiliations

  • Claude Sammut
    • 1
  1. 1.University of New South WalesSydneyAustralia