Fundamentals of the Support Vector Machines for classification. A work for the convex optimization course of das-ufsc.
TeX Matlab
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
code
figures
Readme.md
references.bib
svm_optimization.pdf
svm_optimization.tex

Readme.md

SVM optimization problem

This was a homework done for a convex optimization class and it's objective is to introduce the optimization problem behind the Support Vector Machines for classification tasks.

The tex file was used to give a presentation of the homework and it contains all the mathematical formulation to explain the SVM optimization problem and how it is used to be solved.

The code was written in Matlab and uses the CVX package to solve the discussed convex problem. The example used in the presentation can be run just using the file main.m. One should note some weird results when running the code; the results always prints something about the primal and dual problems, even when I tried to solve just one of them (feel free to fork and make it right).

I made more detailed notes about the subject and I would be happy to share them. If you think they could help, just drop me an email.