Perceptron - Godot Engine

The perceptron, also known as a neuron, is used in machine learning as the building block of a neural network. It is used to classify the results of linerarly separable problems.

It receives several inputs and returns a value at the output.

In this example, the perceptron tries to find the boundary between black and white points in a 2d plane.

See also my projects :

• godot-neural-network
• godot-neuroevolution

How to use

You can instantiate a new perceptron with its number of inputs and its learning rate.

var my_perceptron = Perceptron.new(INPUT_NUMBERS, LEARNING_RATE)

To train the perceptron, pass the input values and the expected result to the training function.

my_perceptron.train([INPUT_1, INPUT_X...], EXEPTED_RESULT)

To get the perceptron result use :

var result = my_perceptron.compute([INPUT_1, INPUT_X...])

Dependencies

This project require Godot Engine 3.3.

Licenses

• The source code is available under the MIT license.