Updated 2015 March 24
This repository hosts code examples from Burak Kanber's Machine Learning in JavaScript blog series at http://burakkanber.com/blog. The series teaches the fundamentals of machine learning algorithms in an accessible manner (no science degree necessary!) through a common, easy language that most web developers know: javascript.
This series targets web developers and individuals withOUT a rigorous computer science background. As such, I generally avoid discussions of linear algebra or higher maths. Algorithms are written procedurally rather than functionally, and I also avoid using linear algebra in the algorithms themselves. It's my belief that machine learning topics are intuitive enough to be understood by most developers, but still the topic is made to feel inaccessible by its reliance on linear algebra and other academic fields that many developers may not have experience in (especially self-taught developers).
Read the introductory blog post here: http://burakkanber.com/blog/machine-learning-in-other-languages-introduction/
This repository simply rehosts the javascript examples. Please note that many of the examples were written in 2012 or earlier, and still need updating in terms of good modern JS practices. And since these examples were first hosted on JSFiddle.net, the source isn't well-organized. I hope to, over time, clean up this repository and modernize the code.
Each directory represents a machine learning topic. Topics with only one
example will contain that example in {folder}/index.html, while topics with
multiple examples will split them up into files named part1.html, etc.
All assets required to run the examples, with the exception of CDN-hosted jQuery, are in the folders as well.
Each folder contains a README.md file, which briefly describes the topic and
provides a link to the blog article(s) that accompany each topic. For now, the
articles are hosted on the blog. Eventually I'll convert them to markdown and
store them in this repository too.
I don't have a Github Pages site set up for this yet, so for now clone the
repository and use your favorite web server to serve this directory. My
favorite is $ python -m SimpleHTTPServer.
- Write README.md's for each folder.
- Modernize the examples.
- Set up a GitHub Pages site for this repository.
- Convert blog posts to markdown and provide them in this repository.
All code in this repository is free for educational use but prohibited for commercial use or any other purpose.
License Notes:
If you would like to use one of these algorithms, write your own implementation! That's the whole point of this series. I want you to learn the algorithms presented, and that means writing your own implementation of them.
And if you need to use any of these algorithms in production, please use a well-known, well-maintained open source project. This code isn't maintained, and it's not optimized for mission-critical use.
That's why this license is so restrictive. Not because I don't want you to use this code, but because you shouldn't.