How to use the steepest descent method to solve a function.
8 views (last 30 days)
Show older comments
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/148746/image.png)
I have an example but I still am not sure how to solve this problem. Please show me step by step on how to attack this. Thank you.
0 Comments
Answers (2)
Vinod Sudheesh
on 26 Jun 2015
Edited: Vinod Sudheesh
on 26 Jun 2015
Hello,
The below code snippet solves this problem using the "Gradient Descend Algorithm"
clc;
clear;
f=@(x)(25*x(1)*x(1)+20*x(2)*x(2)-2*x(1)-x(2));
x=[3 1]';
gf=@(x)([(50*x(1)-2) ; (40*x(1)-1)]);
n=1;
while(norm( gf(x))>0.05)
x= x-0.01*(1/n) *gf(x);
n=n+1;
end
Note that to solve this problem using the "Steepest Descend Algorithm", you will have to write additional logic for choosing the step size in every iteration.
1 Comment
Sarvjeet Singh
on 28 Sep 2020
f=@ (x,y) a*(exp(bx))+c*x
how to solve
determine the general parameters for the non-linear fit using steepest descent method if the fit is given by for the data: ![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/367486/image.png)
![](https://www.mathworks.com/matlabcentral/answers/uploaded_files/367486/image.png)
See Also
Categories
Find more on Least Squares in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!