In supervised learning, example inputs are given to the computer and labelled with the desired outcomes. The idea behind this approach is to enable the algorithm to "learn" by comparing its real output with the "learned" outputs in order to identify faults and adjust the model as necessary. Therefore, supervised learning makes use of patterns to forecast label values on new unlabeled data.
In supervised learning, for instance, an algorithm might be fed data that includes pictures of sharks categorised as fish and pictures of oceans labelled as water. The supervised learning algorithm ought to be able to recognise unlabeled shark pictures as fish and unlabeled ocean pictures as water after being trained on this data.
Using previous data to forecast future events that are statistically likely is a frequent application of supervised learning. It might be used to filter out spam emails or to forecast future stock market changes using past data. Untagged images of dogs can be classified using supervised learning techniques using tagged images as input data.