This post is an introduction to gradient descent, which is a powerful tool for optimization. It looks at both the theory as well as an implementation.