Machine Learning in the Browser with Deep Neural Networks
summary
Shows a working JavaScript implementation of neural nets in the browser: a naive perceptron demo that learns a linear separator, experiments in the TensorFlow Playground with hidden layers that fix XOR‑like patterns, and a run of a pre-trained convolutional net from Google that classifies images. Reports training vs test loss, demonstrates overfitting, shows accuracy dropping from random to under 0.1% training loss on a toy dataset, and emphasizes zero-install deployment and CPU/GPU cost tradeoffs.
AI-written summary · about
chapters (11)
- 0:00Start
- 0:29What Is Machine Learning
- 3:46Deep Neural Networks
- 7:40Interactive Learning
- 9:36Tensor Flow Playground
- 9:54Classification Example
- 12:04Introduction
- 12:54Hidden Layer
- 15:10Stochastic Gradient Descent
- 16:11Training Loss
- 22:14Convolutional Networks
related talks
29mMachine Learning with Deep Neural Networks in the Browser
36mMachine Learning in the Browser
35mThe Future of JavaScript & Machine Learning
48mDeep Learning in the Browser
32mDeep Learning in JS
39mRun Deep Learning models in the browser with JavaScript and ConvNetJS
51mMachine Learning Exposed Deep and Reinforcement Learning
35mMachine Learning in the Browser with deeplearnJS