Journey to the center of the perceptron

Building a perceptron from scratch has been the most satisfying experience

This page is my attempt to unify my understanding of perceptrons. It’s a mix between tutorial, explanation, and history.

Caution

Read this as a beginner

This page is my attempt to understand perceptrons.

Motivation

• A logical calculus of the ideas immanent in nervous activity

• The perceptron: A probabilistic model for information storage and organization in the brain

• Perceptron

• Building a Perceptron from Scratch

• Shallow Neural Network in Keras

import math
import random

import numpy as np


class Perceptron:
    def __init__(self, input_size=2, learning_rate=0.01, epochs=5):
        self.learning_rate = learning_rate
        self.epochs = epochs
        self.input_size = input_size
        self.weights = np.random.uniform(-1, 1, size=(input_size))
        self.bias = random.uniform(-1, 1)

    def sigmoid(self, X):
        return 1 / (1 + math.exp(-X))

    def z(self, X):
        return np.dot(X, self.weights) + self.bias

    def predict(self, X):
        z = self.sigmoid(np.dot(X, self.weights) + self.bias)
        return 1 if z > 0.5 else 0

    def fit(self, X, y):
        for n in range(self.epochs):
            for x, y_actual in zip(X, y):
                y_predicted = self.predict(x)
                error = y_actual - y_predicted
                self.weights += self.learning_rate * error * np.array(x)
                self.bias += self.learning_rate * error


X = [[0, 0], [0, 1], [1, 0], [1, 1]]
y = [0, 1, 1, 1]
perceptron = Perceptron()
perceptron.fit(X, y)