Training Neural Networks in Python [Repost]

Prerequisites

You must feel comfortable writing code in Python 3.

Project

Build a neural network from scratch using Python.

Neural networks are widely used in everyday applications like online shopping, weather forecasting, and smartphones. This course by Eduardo Corpeño teaches the inner workings of neural networks in Python, allowing you to fully understand the algorithms behind them. Although there are professional tools that allow you to train neural networks from a high-level perspective, this course gives you a chance to tap into the details of the algorithms behind neural networks. Through exercises and examples, learn how to relate biological neurons to Python elements to build and train your own networks, and gain knowledge that can help you choose the right neural network architecture and training method for your projects and problems you encounter.

This course is integrated with GitHub Codespaces, an instant cloud developer environment that offers all the functionality of your favorite IDE without the need for any local machine setup. With GitHub Codespaces, you can get hands-on practice from any machine, at any time—all while using a tool that you’ll likely encounter in the workplace. Check out the “Using GitHub Codespaces with this course” video to learn how to get started.

Learning objectives

• Understand the difference between neural networks and other programming tools.
• Determine whether a neural network is the best choice for a specific problem in machine learning.
• Learn how to add different types of neural networks to your toolbox for solving problems other than classification.
• Design a multilayer perceptron graphically from a set of parameters like the number of inputs, outputs, and layers.
• Relate parts of a biological neuron to Python elements to create a model of the brain.
• Create a simple model for a neuron based on a weighted sum and choose from a variety of activation functions to solve the problem of different neuron output sizes.
• Create a perceptron model, which is the final building block of a neural network.