Train the machine using data which is well labeled. For example we need to train the machines to identify vegetables with identification(colors and shapes). Based on that machine can identify vegetables. Here target is very clear.Supervised learning has two categories
Target/response variables take only descrete(finite/countable) values
Target is a continuous variable/can take any real number
There is no data to identify the things. Machine has to defferentiate/categoroze the things which are available. For example we have given 4 variaties of dogs and three cats. System is not having data to recognize the dog and cat. But it will identify 4 are same category and 3 are belongs to another category.
Technique to represent a forward step for machine learning. Is like how human brain works.