Implementation of the steepest descent algorithm in MATLAB from scratch.

Goal is to find the local minimum (if it exists) using non-linear algorithms such as Steepest Descent.

Example:

Reviewing this function, we conclude that it is in normal form.

With initial guess=[0.5; 0.5] and step alpha=0.15 (decided from grid search), calculated within 138 repetitions, f has a local minimum at x*=[1; 1].

As we can see in the diagram, the method is quite slow and fluctuates until it finally reaches the minimizer [1,1].