Neural network design sprint for Phaser game "Dodge Lasers" скачать в хорошем качестве

Neural network design sprint for Phaser game "Dodge Lasers"

Timestamps 0:00 – MrMultiMediator 0:05 – Introductory remarks and overview. 1:12 – Neural network explained broadly. 2:30 – Layers, and associated indexed quantities explained (which neurons do they correspond to? Which neurons do they connect and how?). Math introduced. 7:45 – Math expanded. How are activations evaluated? Sigmoid function explained. Significance of the bias explained. 10:25 – Intended implementation for ‘Dodge Lasers’ game explained. 11:55 – Introduction to matrix multiplication 14:18 – Matrix multiplication example in JavaScript with math.js library 16:55 – Connection between matrix multiplication to deep learning/neural nets explained. This is a brief overview of how neural networks work as well as a quick design session for the neural network and how it will be implemented in the ‘Dodge Lasers’ game. There is also an intro to matrix multiplication and an example implementation of matrix multiplication in JavaScript using the math.js library. Finally, the relevance of matrix multiplication for deep learning and neural networks is explained. Outro music is on soundcloud:   / theme-coup-de-cast   Previous videos in this series: How to run JavaScript in Chrome:    • How to run JavaScript in Browser   Code tutorial – Dodge Lasers (JavaScript game) walkthrough -    • Code tutorial - Dodge Lasers (JavaScript g...