WHAT PERCEPTRON IS?

perceptron is nothing but a mathematical setup that takes certain  attributes from data and multiply the attributes with weights  and then apply summation to these products. you will get a linear expression. after all pass the expressions calculated value to activation  function (like sigmoid function or step, function). you will get some probablisṭic value by sigmoid function (i.e between [-1,1]) and we  can classify the data on the basis of this value into two classes.

WHERE PERCEPTRON IS USED?

Well ,Perceptron Is Basic  Building Block Of  "Nural Networks".
And Nural Network Is The Most Basic Algoritham Used In Deep Learning.

Below Picture Illustrate What A Nural Network Can Do.

Here ,It Is Taking Different Pictures As Input And Identify What Animal Is This.

There Exist Various Types Of Nural Network ,Below Listed Are Some Among Them:

Multi-layer Perceptrons (MLPs)

• Basic neural network
• Used for simple tasks

Convolution Neural Networks (CNNs)

• Uses filters and pooling to find characteristics in data
• Mostly used for image tasks

Recurrent Neural Networks (RNNs)

• Uses the previous result of data in figuring out new output
• Used for temporal data