Perceptron

Recall the classification of points problem in the previous lesson. How do we draw the line that separates the squares from the stars? This is where perceptron comes to the rescue.

What is a perceptron?

Perceptron is a binary linear classifier (binary) used in supervised learning to classify the given input data.

Visualize a perceptron

In case of two input features the perceptron model would look something like this:

Boundary

The boundary line that separates the two classes is:

$w_1x_1 + w_2x_2 + b = 0$

Here,

• $x_1$ and $x_2$ are the inputs
• $w_1$ and $w_2$ are the weights
• $b$ is the bias

Prediction

$\hat{y}$ = { 1 if $w_1x_1 + w_2x_2 + b$ ...