Real Time Machine Learning With Node js
summary
Claims Node.js can do real‑time, incremental ML using actual packages and data streams rather than just rhetoric: a custom numeric library lip, streaming from CouchDB/npm changes, Cheerio for scraping, and simple modules for incremental mean and moving mean power live stats. An online regression on package age shows a positive slope (roughly two more versions per year in the demo), and a classifier flags webpack as the strongest predictor of React packages via confusion matrices and feature importances.
AI-written summary · about
chapters (12)
- 0:00Intro to real-time ML with Node.js
- 1:30Why JavaScript for ML
- 3:38Real-time vs batch ML
- 5:01Incremental mean demo
- 7:11Non-stationary data challenges
- 8:37Moving average for trends
- 10:02Three ML problem types
- 13:34Regression with npm
- 16:25Binary classification of React packages
- 21:00Clustering methods
- 22:02Recommended machine learning resources
- 23:03Closing & Q&A
related talks
32mUsing Machine Learning Models in JavaScript Application- 32mUsing Machine Learning Models in JavaScript Application
01h 05mIntro to machine learning in Node
54mIntroduction to machine learning in Node.js
15mThe future of machine learning using JavaScript
39mMachine Learning Magic for Your JavaScript Application
23mThe new era of ML in JavaScript
18mMachine Learning with Javascript