Contractions takes an objective function and uses its two-argument output to find a fixed point. This algorithm in particular can
1) handle the routine in the background while parameters are easily manipulated.
2) can accelerate and dampen stepsizes while it recognizes when to dampen it
3) it respects box constraints.
4) functional arguments from the objective functions can be passed on to carry over "good" starting values for problems in the objective.
5) it can switch to a levenberg-marquardt algorithm. The LM algorithm can use the Broyden method to update the Jacobian which reduces function evaluations.
The LevenbergMarquardt needs to be downloaded for full functionality:
http://www.mathworks.com/matlabcentral/fileexchange/53449-levenberg-marquardt-toolbox
The Jacobian toolbox is also required for the LevenbergMarquardt algorithm:
http://www.mathworks.com/matlabcentral/fileexchange/53440-jacobian-toolbox
인용 양식
Alexander (2024). Contractions - algorithm and example for nested fixed points (https://www.mathworks.com/matlabcentral/fileexchange/53451-contractions-algorithm-and-example-for-nested-fixed-points), MATLAB Central File Exchange. 검색됨 .
MATLAB 릴리스 호환 정보
플랫폼 호환성
Windows macOS Linux카테고리
- Image Processing and Computer Vision > Image Processing Toolbox > Image Segmentation and Analysis > Region and Image Properties >
태그
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!버전 | 게시됨 | 릴리스 정보 | |
---|---|---|---|
3.0 |
|