TOPICS
Search

Bayes Classifier


A Bayes classifier is a classification rule that applies the Bayes decision rule. Given probabilities P(C_k|x) under a posterior distribution for classes C_1, ..., C_K and a loss function L(a,C_k), it chooses an action a that minimizes the posterior expected loss

 R(a|x)=sum_(k=1)^KL(a,C_k)P(C_k|x).

For a zero-one loss function, this is equivalent to choosing the class having largest such probability,

 C^^(x)=argmax_(C_k)P(C_k|x).

See also

Bayes Decision Rule, Classification, Loss Function, Pattern Recognition, Posterior Distribution

Explore with Wolfram|Alpha

References

Bishop, C. M. Pattern Recognition and Machine Learning. New York: Springer, 2006.

Cite this as:

Weisstein, Eric W. "Bayes Classifier." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/BayesClassifier.html

Subject classifications