Welcome to Chapter 10 of The Nature of Code: Neural Networks. In this video, I provide a brief introduction neural networks and an overview of topics in upcoming videos.

In this series, I build a “doodle classifier” using the Google “Quick, Draw!” dataset and my JavaScript neural network library.

In this video, I continue my machine learning series and build a simple Perceptron in Processing (Java).

In this “doodle classifier” video, I cover how to parse the Google Quick Draw dataset with Processing and prepare it for use with the JavaScript neural network library.

This is a follow-up to my Perceptron Video (https://youtu.be/ntKn5TPHHAk)

In part 3 of the “doodle classifier” series, I use the Google Quick Draw data prepared in Processing and import it into a p5.js JavaScript sketch.

In this video, I move beyond the Simple Perceptron and discuss what happens when you build multiple layers of interconnected perceptrons (“fully-connected network”) for machine learning.

In this video, I create a neural network object and train it with a subset of the Google Quick Draw dataset.

This video follows up on the previous Multilayer Perceptron video. Here I begin the long process of coding a simple neural network library in JavaScript.

In this video I “test” the neural network’s classification accuracy with drawings from Google Quick Draw dataset not used for training.

In this video, I introduce the idea of “Linear Algebra” and explore the matrix math required for a simple neural network library.

In this video, I use the p5.js library to create a drawing canvas for the user and attempt to classify doodles using the trained neural network model.

This video is a continuation of my first Matrix Math video where I implement in JavaScript the math operations that I need in order to create my own simple neural network library.

In this video, I update the matrix.js file to use ES6 classes.

This video continues the “Matrix Math” series where I attempt to implement in JavaScript the math operations required for a simple neural network library. In this 3rd part, I implement matrix multiplication (“matrix product”).

This video continues the “Matrix Math” series where I attempt to implement in JavaScript the math operations required for a simple neural network library. In this 4th part, I implement a transpose() function for a matrix.

In this video, I make small improvements to the Matrix class for th upcoming neural network example. Those improvements are: static methods, a “print” method, and a “map” method.

In this video, I tackle a fundamental algorithm for neural networks: Feedforward. I discuss how the algorithm works in a Multi-layered Perceptron and connect the algorithm with the matrix math from previous videos.

This video is a continuation of the Feedforward algorithm video. In this part, I implement the code for the algorithm in a NeuralNetwork class written in JavaScript.

In this video, I discuss the backpropagation algorithm as it relates to supervised learning and neural networks.

In this video, I begin implementing the backpropagation algorithm in my simple JavaScript neural network library.

In this video, I discuss how “gradient descent” can be used to adjust the weights during back propagation in my “toy” JavaScript neural network library.

In this video, I implement the formulas for “gradient descent” and adjust the weights in the train() function of my “toy” JavaScript neural network library.

In this video, I implement the formulas for “gradient descent” and adjust the bias in the train() function of my “toy” JavaScript neural network library. I also test the library with a simple XOR dataset.