User:Samohyl Jan/Bayesian filtering

From Wikipedia, the free encyclopedia

This article is about statistical method. For application of Bayesian methods in classification of text or spam, see Bayesian spam filtering.

Bayesian filtering, is a generic name of statistical methods for obtaining hidden state of dynamical system from its observables and history using bayesian inference. It is a generalization of well known hidden Markov models and Kalman filters.

Contents 1 Basic idea 2 See also 3 External links 4 TODO!!!

[edit] Basic idea

One is given:

[edit] See also

• hidden Markov model
• Kalman filter

[edit] External links

A survey of probabilistic models, using the Bayesian Programming methodology as a unifying framework

[edit] TODO!!!

• Move Bayesian filtering to Bayesian spam filtering
  • fix intro
  • mention naive Bayes classifier
• Describe the model
• Note that we don't require dynamical system to be deterministic
• fix link to Bayesian filtering at Project Cybersyn and check other places