Classification with TensorFlow

In Chapter 1, Introduction to Machine Learning, we talked about supervised learning and briefly talked about classification modeling. Classification modeling involves predicting classes in our target variable. When the classes we try to predict are binary (for example, trying to predict whether a pet is either a dog or a cat, whether an email is spam or not, or whether a patient has cancer or not), this type of classification scenario is referred to as binary classification.

Then again, we may be faced with a problem where we want to build an ML model to predict the different breeds of dogs. In this case, we have more than two classes, so this type of classification is called multi-class classification. Just like binary classification problems, in multi-class classification, our target variable can only belong to one class out of multiple classes – our model will select either a bulldog, a German shepherd, or a pit bull. Here, the classes are...