Skip to content

Latest commit

 

History

History
39 lines (23 loc) · 804 Bytes

README.md

File metadata and controls

39 lines (23 loc) · 804 Bytes

colint

Collective intelligence library

purpose

colint defines a set of classes and functions useful for collective intelligence / artificial intelligence tasks. (classification, regression, estimation etc...)

This repository was inspired by the book Collective Intelligence by Toby Segaran 2007.

example

colint = require 'colint'

classify = colint.classify

# create a NaiveBayes classifier with a simple token splitter
nb = new classify.NaiveBayes(classify.splitters.simple)

nb.train "This is the text of some document", "category"

[top_category, stats] = nb.classifications "Document to classify"

see /examples folder for a working classification, recommendation and decision tree example

install

With npm do:

npm install colint

license

MIT