IN THIS CHAPTER
Understanding how machine learning works under the hood
Recognizing the different parts of a learning process
Defining the most common error functions
Deciding which error function is the best for your problem
Glancing at the steps in machine learning optimization when using gradient descent
Distinguishing between batch, mini-batch, and online learning
Machine learning may appear as a kind of magic trick to any newcomer to the discipline — something to expect from any application of advanced scientific discovery, as Arthur C. Clarke, the futurist and author of popular sci-fi stories (one of which became the landmark movie 2001: A Space Odyssey), expressed by his third law, stating, “any sufficiently advanced technology is indistinguishable from magic.” However, machine learning isn’t magic at all. It’s the application of mathematical formulations to how we view the human learning process.
Expecting that the world itself is a representation of mathematical and statistical formulations, machine learning algorithms strive to learn about such formulations by tracking them back from a limited number of observations. Just as you don’t need to see all the trees in the world to learn to recognize one (because humans can understand the distinguishing characteristics of trees), so machine learning algorithms can use the computational power of computers and the wide availability of data about everything to learn how to solve a large ...