Learn Neural Networks through coding

In this video I teach you what neural networks are, how they work and how to implement them using JavaScript. We will experiment by using the self-driving car and having it make its own decisions.

Neural networks are computing systems inspired by the biological neural networks in our brains. A neuron is a specialized cell that communicates with other cells via synapses. Branch-like structures (the dendrites) receive the signals. When stimulated enough, a neuron will fire a signal through its axon. So, a single neuron does something very simple and intelligence only comes when you have many of them working in a team.

Artificial neural networks work so that neurons on the first layer will be connected to the sensors. They will send signals forward a few times times... finally, the last layer will be connected to the car controls to make it actually do something.


⭐️MORE LINKS⭐️
3b1b video: https://youtu.be/aircAruvnKk
Plane intersection visualizer: http://www.songho.ca/math/plane/plane.html


⭐️CODE⭐️
Github: https://github.com/gniziemazity/Self-driving-car (6. Neural network)
Use '5. Traffic simulation if you plan to follow along!
My website: https://radufromfinland.com
JS NN Libraries: https://blog.logrocket.com/best-javascript-machine-learning-libraries-in-2021


⭐️ TABLE OF CONTENT ️⭐️
0:00 Intro
0:58 What are neural networks
4:11 Implementing a neural network in JavaScript
8:35 Feedforward algorithm
10:32 Mathematics of neural networks
13:30 Hidden layers
16:09 Connecting the network to the self-driving car
22:12 Debugging a neural network