Programming Machine Learning

What You Just Learned

In this chapter you learned about a powerful concept: the decision boundary of a classifier. During the training phase, a classifier shapes the decision boundary to separate the classes in the dataset. Thereafter, it classifies data points based on which side of the boundary they fall.

Now that you know about the decision boundary, you also understand why a neural network is generally more powerful than a perceptron. A perceptron can only draw a straight decision boundary, so it only works well with linearly separable data. A neural network is happy to bend its decision boundary around pretzeled data.

Now that you have a better appreciation for neural networks and how they work, you might wonder: if neural networks are so ...