Naive Bayes Classifier

Classifies continuous or binary data using a naive bayes strategy in Node.js.

None of this is vectorized, so it's not very fast.

#Dependencies Requires Node.js. Download it from nodejs.org.

I use stats-lite, gaussian, underscore, and promise libraries that can be downloaded from npm. They are all listed in the package.json file, so just run npm install to get all the necessary libraries.

#API dataType can be 'binary' or 'continouous_value'
X is a 2-d array
y is a 1-d array

#Example

var Fitter = require('./fitter');
var Predictor = require('./predictor');

var trainingX2 = [[1,0],[0,1],[1,0],[0,1],[1,1],[0,1]];
var trainingY2 = ['Sydney','Paris','Sydney','Paris','Sydney','Paris'];
var testX2 =[[1,0],[0,1],[1,1]];
var dataType = 'binary'

var fitter = Fitter(trainingX2,trainingY2,dataType);

fitter.fit(function(fitter){
	var predictor = Predictor(fitter,testX2);
	predictor.predict(function(y){
		console.log(y);
	});
});