 |
|
|
|
|
 |
 |
|
Mixture Estimation using the EM Algorithm
|
|
|
|
|
|
|
|
|
|
|
|
0207-683
|
|
|
|
|
|
1995-05-29
|
|
|
|
|
|
This package uses the EM (expectation maximization) algorithm to simultaneously estimate the missing data and unknown parameter(s) associated with a data set. Here the missing data are assumed to be the identities of the observations originating from each of the two distributions contributing to the mixture. The parameters describe the component distributions of the mixture; the distributions may be continuous or discrete.
A significance test comparing the likelihood of the data arising from the mixture distribution versus the likelihood of the data arising from the first component distribution is provided in the output. Examples are included at the end of the package.
|
|
|
|
|
|
|
|
|
|
|
|
expectation maximization (EM), parameter estimation, maximum likelihood estimation, mixture distribution
|
|
|
 |
|
|
| MixtureEstimation.m (36.9 KB) - Mathematica Package |
|
|
For the newest resources, visit Wolfram Repositories and Archives »
