Im trying to use gradient decent with a for loop to determine the value of x that minimizes the function

5 views (last 30 days)
Im trying to use gradient descent to numerically determine the value of x that
minimizes the function f(x) = x^2 − 3x + 1 with its derivative f(x) = 2x − 3.
Starting from x(0) = 0, im trying to plot the value of the
function f(x) as a function of the number of iterations for a few different
values of α.
This is my current code:
clc, clearvars, close all
iterations = 10;
alpha = 0.15;
x = zeros(iterations,1);
x(1) = -1;
for k = 2:iterations
x(k) = x(k-1) - alpha*( 2*x(k)-3 )
end
figure(1)
plot(k,x)
figure(2)
plot(x1,f_x)
Im not sure how to plot this, after its calculated i could use f= x.^2-3*x+1 to calculate its function value.
Any ideas how i could make it work?

Answers (2)

David Hill
David Hill on 22 Nov 2022
iterations = 10;
alpha = 0.15;
x = zeros(iterations,1);
x(1) = -1;
for k = 2:iterations
x(k) = (x(k-1)+alpha*3)/(1+2*alpha);
end
plot(x)

Torsten
Torsten on 22 Nov 2022
Edited: Torsten on 22 Nov 2022
f = @(x) x.^2 - 3*x + 1;
iterations = 15;
alpha = 0.15;
x = zeros(iterations,1);
x(1) = 0;
for k = 2:iterations
x(k) = x(k-1) - alpha*( 2*x(k-1)-3 ) ;
end
figure(1)
hold on
plot(x,f(x),'o')
xplot = linspace(0,2,100);
yplot = f(xplot);
plot(xplot,f(xplot))
hold off
grid on
figure(2)
plot(1:iterations,x,'o')
grid on

Categories

Find more on Interactive Control and Callbacks 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!