Add a genetic algorithm as a minimizer
Registered by
Tom Dimiduk
Genetic algorithms (and other stochastic fitting methods) are supposed to be better at fitting for large numbers of parameters. They are also less dependent on good initial guesses and more able to avoid local minima.
This is probably becoming even more important as we move to fitting ever larger objects.
[VNM]: we should get a genetic algorithm for free with OpenOpt, once we implement the alternate-minimizer blueprint.
Blueprint information
- Status:
- Not started
- Approver:
- None
- Priority:
- Medium
- Drafter:
- None
- Direction:
- Approved
- Assignee:
- Rebecca W. Perry
- Definition:
- New
- Series goal:
- None
- Implementation:
- Unknown
- Milestone target:
- future
- Started by
- Completed by
Related branches
Related bugs
Sprints
Whiteboard
(?)
Work Items
Dependency tree
* Blueprints in grey have been implemented.