Ockham's Razor is the philosophical principle that entities or assumptions should not be introduced unnecessarily. This is often interpreted as stating that one should seek the simplest explanation for a phenomenon. The purposes of this note are to introduce some of the different formulations of the Razor and the controversy surrounding it, to explore the use of Ockham's Razor as a scientific tool, and to draw connections with statistical inference, both frequentist and Bayesian. The note starts with a brief introduction to the life of William of Ockham, to whom the Razor is traditionally and popularly attributed. There is some controversy regarding the use of Ockham's Razor as an inferential tool; this seems to center on the meaning of ‘simple explanation’ and the issue is explored. Uses of the Razor in science, in statistical model selection, and in Bayesian inference are examined. Copyright © 2010 John Wiley & Sons, Inc.
For further resources related to this article, please visit the WIREs website.
[1]
Vijay Balasubramanian,et al.
Statistical Inference, Occam's Razor, and Statistical Mechanics on the Space of Probability Distributions
,
1996,
Neural Computation.
[2]
R. Lewis,et al.
Statistical models and Occam's razor.
,
1999,
Academic emergency medicine : official journal of the Society for Academic Emergency Medicine.
[3]
R. Tibshirani.
Regression Shrinkage and Selection via the Lasso
,
1996
.
[4]
Stephen M. Stigler,et al.
STIGLER'S LAW OF EPONYMY†
,
1980
.
[5]
C. Robert,et al.
Harold Jeffreys’s Theory of Probability Revisited
,
2008,
0804.3173.
[6]
W. M. Thorburn,et al.
THE MYTH OF OCCAM'S RAZOR
,
1918
.
[7]
D. Spiegelhalter,et al.
Bayes Factors and Choice Criteria for Linear Models
,
1980
.
[8]
G. Schwarz.
Estimating the Dimension of a Model
,
1978
.