Machine Learning Math: Understanding Classification Problems

In our previous blog posts, we covered regression problems where we’re trying to predict the value of our target function. In regression problems, our target function is a real-valued function, and we’re trying to compute a real number that estimates its value at given data points.

In the next couple of posts, we’ll cover classification problems, which are also known as selection problems. In a classification problem, our target function is returning one item from a discrete set of choices: in other words, it’s making a “true or false” decision or picking one option out of a few. Such a function might look at an image and determine whether the image contains a person’s face, or it could look at a hand-written letter or number and identify the character.