Difference between revisions of "Inductive bias"
From Lesswrongwiki
(fleshed out) |
|||
Line 12: | Line 12: | ||
}}</ref>}} | }}</ref>}} | ||
− | + | "Inductive bias" refers to your suspicion that if the sun has risen for the last billion days in a row, then it may rise tomorrow as well. Since it is [[logically possible]] that the laws of physics will arbitrarily cease to work and that the sun will *not* rise tomorrow, coming to this conclusion requires an inductively biased [[prior]]. | |
− | ==See also== | + | This sort of bias is not a bad thing - without "inductive bias" you can't draw any conclusion at all from the data. It's just a different technical meaning attached to the same word. |
+ | |||
+ | ===Primary article=== | ||
+ | *[http://lesswrong.com/lw/hg/inductive_bias/ "Inductive Bias"] by [[Eliezer Yudkowsky]] | ||
+ | |||
+ | ===See also=== | ||
*[[Prior distribution]] | *[[Prior distribution]] | ||
*[[Statistical bias]] | *[[Statistical bias]] | ||
+ | *[[Cognitive bias]] | ||
− | ==Footnotes== | + | ===Footnotes=== |
<references/> | <references/> | ||
− | |||
− | |||
− | |||
− | |||
− |
Revision as of 09:57, 5 June 2009
The inductive bias of a learning algorithm is the set of assumptions that the learner uses to predict outputs given inputs that it has not encountered
—Tom Mitchell[1]
"Inductive bias" refers to your suspicion that if the sun has risen for the last billion days in a row, then it may rise tomorrow as well. Since it is logically possible that the laws of physics will arbitrarily cease to work and that the sun will *not* rise tomorrow, coming to this conclusion requires an inductively biased prior.
This sort of bias is not a bad thing - without "inductive bias" you can't draw any conclusion at all from the data. It's just a different technical meaning attached to the same word.
Primary article
See also
Footnotes
- ↑ Tom M. Mitchell (1980) (PDF). The need for biases in learning generalizations. http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.120.4179.