Udemy
    •  
    •  
    •  
    •  
    •  
    •  
    •  
    •  
Subscribe
Turn what you know into an opportunity and reach millions around the world.
Learn More
Your cart is empty.
Keep shopping
Machine Learning: Build neural networks in 77 lines of code
Highest Rated
Rating: 4.8 out of 5(959 ratings)
3,141 students

Machine Learning: Build neural networks in 77 lines of code

Machine Learning and Artificial Intelligence for beginners. How to build a neural network in 77 lines of Python code.
Highest Rated
Created byMilo Spencer-Harper
Last updated 3/2019
English

What you'll learn

  • Neural Networks
  • Machine Learning
  • Artificial Intelligence
  • Supervised Learning

Course content

2 sections • 18 lectures • 56m total length

  • Course Structure1:04

    In this section, I briefly outline the structure of the course.

  • What is a neural network?1:18

    In this section, I'll explain how a biological neural network works, which in turn has inspired artificial neural networks which run on computers.

  • The challenge1:42

    In this section, I will show you the problem our neural network is going to solve.

  • Designing our architecture0:28

    In this section, I will show you how to design the architecture of a neural network to fit the problem.

  • Weights1:27

    In this section, I will introduce you to the concepts of weights. Each input to a neuron has a weight, which indicates how much the signal should be amplified or reduced. The weights act as memory.

  • Activation Function3:46

    In this section, I will introduce you to the activation function. What is an activation function? An activation function is a mathematical formula which describes how the output of a neuron varies as its input changes. There are many different formulas which could be used. In this video, we will use the sigmoid formula. Watch the video to learn more.

  • Training process1:42

    In this section, I will show you how the neural network uses the training set to learn.

  • Error Cost Function1:22

    In this section, I will introduce you to the Error Cost Function. This describes the accuracy of the neural network. Although we don't use an Error Cost Function in our Python code directly, it gives you the foundational knowledge to understand the next section.

  • Adjusting the Weights4:07

    In this section, I explain gradient descent and we derive the the critical formula for adjusting the weights. This formula is the final piece in the puzzle, that will allow us to build our neural network.

Requirements

  • Basic Python knowledge

Description

From Google Translate to Netflix recommendations, neural networks are increasingly being used in our everyday lives. One day neural networks may operate self driving cars or even reach the level of artificial consciousness. As the machine learning revolution grows, demand for machine learning engineers grows with it. Machine learning is a lucrative field to develop your career.


In this course, I will teach you how to build a neural network from scratch in 77 lines of Python code. Unlike other courses, we won't be using machine learning libraries, which means you will gain a unique level of insight into how neural networks actually work. This course is designed for beginners. I don't use complex mathematics and I explain the Python code line by line, so the concepts are explained clearly and simply.


This is the expanded and improved video version of my blog post "How to build a neural network in 9 lines of Python code" which has been read by over 500,0000 students.

Enroll today to start building your neural network.

Who this course is for:

  • Anyone interested in machine learning

Report abuse