Supervised learning

In this type of learning, computers learn from a predefined data set with data labels and features. Its aim is to predict an output based on the given input variables using the defined data point and label from the learned dataset. The most important thing you need in supervised learning is data with labels. By providing data labels, we teach and train our model for accuracy. The more accurate your data is, the closer your prediction will be. Some of the ways of getting data include from a historical source, or by doing experiments.

Now let's take an example of supervised learning. In your mailbox, you have a folder for junk emails; ever wondered how it automatically identifies emails as spam? It is actually based on the trained model, which looks for certain things before marking it as junk, including the source from where the email was generated, the intended audience (whether it is directly targeting the recipient), whether the email body contains marketing or spam...