Perceptron: Fundamental Block of Deep Learning with Python

Perceptron basic and its implementation for artificial intelligence

Artificial Neural Network. A photo by the Author

Introduction

It is an algorithm to solve the problem of supervised machine learning. Way of its structure we can call it a mathematical model and functions. It is also a foundation of the deep learning algorithms architecture.

The diagram of the perceptron is shown below:

A photo by the Author

where,

w1, w2 denotes the weights and b denotes the bias
x1 and x2 are the inputs
summation ( dot product) = w1x1 + w2x2 + b = z
f = activation function

The activation function is a mathematical function to make the value z in some range based on the condition. Examples of activation functions are step function, sigmoid, tanh, relu, etc.

As we do training and prediction in machine learning, the perceptron is also trainable, and based on that we can do prediction. In the training phase, the main objective is to find the optimal value of weights and bias to achieve highly…