Opened 13 years ago
Last modified 13 years ago
#1439 new task
[GrzLinModel] Allowing distributions with extra parameters
Reported by: | Víctor de Buen Remiro | Owned by: | Víctor de Buen Remiro |
---|---|---|---|
Priority: | highest | Milestone: | Mantainance |
Component: | Math | Version: | 3.1 |
Severity: | blocker | Keywords: | |
Cc: |
Description
Under certain circumstances, the optimization and simulation methods used over pure exponential family, can be extended to distributions with extra parameters than average.
This extra parameters can be simulated in a Gibbs frame by simple alternating and incorporating prior information about all them. In MLE optimization we can use Expectation–maximization algorithm to take profit of reusing written code.
Change History (3)
comment:1 Changed 13 years ago by
Priority: | normal → highest |
---|---|
Severity: | normal → blocker |
Version: | 2.0.1 → 3.1 |
comment:2 Changed 13 years ago by
Note: See
TracTickets for help on using
tickets.
(In [4165]) Refs #1439
Fixes #1440