In our classification system, we have data that is used to train our model. In this case of sorting emails into clusters, discrete values are provided with the data, and this is known as classification.
There is another aspect of supervised learning, where instead of providing a discrete value, we provide it with a continuous value. This is known as regression. Regression is also considered supervised learning. The difference between classification and regression is that the first has discrete values and the latter has continuous, numeric values. The following diagram illustrates the three learning algorithms that we can use:
As you can see in the preceding diagram, we use Supervised Learning, Unsupervised Learning, and Reinforcement Learning. When we talk about Supervised Learning, we also use Classification. Within Classification, we perform tasks such as Identify Fraud Detection, Image Classification, Customer Retention, and Diagnostics. In Regression, we perform activities such as Advertising Popularity Prediction, Weather Forecasting, and so on. In Reinforcement, we perform Game AI, Skill Acquisition, and so on. Finally, in Unsupervised Learning, we have Recommender Systems and different sub-fields of machine learning, as illustrated.
