Difference between revisions of "Parameter estimation"
Jump to navigation
Jump to search
m |
|||
| Line 1: | Line 1: | ||
| − | In probability, once a | + | In probability, once a model class has been selected it is then necessary to estimate the values of its associated ''parameters'' in order to specify that particular model out of all members of the model class which differ only on the basis of the values of these parameters. |
| − | In [[Bayesian Probability]], parameters are estimated by computing the joint [[posterior distribution]] of the | + | In [[Bayesian Probability]], the parameters are estimated from data by computing the joint [[posterior distribution]] of the parameters and the data using [[Bayes Rule]]. |
[[category:mathematics]] | [[category:mathematics]] | ||
Revision as of 23:28, February 27, 2009
In probability, once a model class has been selected it is then necessary to estimate the values of its associated parameters in order to specify that particular model out of all members of the model class which differ only on the basis of the values of these parameters.
In Bayesian Probability, the parameters are estimated from data by computing the joint posterior distribution of the parameters and the data using Bayes Rule.