fagan2888 / rudy.js

A ML library for JavaScript (in development)

rudy.js - A ML library for JavaScript (in development)

(Rudy: cause it might not be fun to do ML in JavaScript, but we just have to keep at it)

###Example data Some data for illustration

var test_set = [{"id": 1, "target": 1, "input": 1, "input2": 2},
             {"id": 2, "target": 1, "input": 2, "input2": 3},
             {"id": 3, "target": 1, "input": 3, "input2": 4},
             {"id": 4, "target": 1, "input": 3, "input2": 4},
             {"id": 5, "target": 0, "input": 5, "input2": 5},
             {"id": 6, "target": 0, "input": 6, "input2": 6},
             {"id": 7, "target": 0, "input": 7, "input2": 7},
             {"id": 8, "target": 0, "input": 5, "input2": 8},
             {"id": 9, "target": 0, "input": 4, "input2": 3},
             {"id": 10, "target": 0, "input": 3, "input2": 5},
             {"id": 10, "target": 1, "input": 3, "input2": 6},
             {"id": 10, "target": 1, "input": 2, "input2": 8},
             {"id": 11, "target": 0, "input": 7, "input2": 9},
             {"id": 12, "target": 0, "input": 6, "input2": 0},
             {"id": 13, "target": 0, "input": 7, "input2": 4}
            ];

##Basic functionality - basics.js ###tab() - basic cross tabs

• v1: given an array and user-specified attributes, returns an array of cross-tabs.

tab(test_set,"input","target")

###summary() - in-development

###unique() - unique values in list

##Supervised Learning - rudyTree.js ###binaryTree() - for discrete tree learning v1: A easy to use decision tree for binary targets and numeric input variables. v2: In development will support discrete input variables and provide facility for ROC curves

inputs = ["input","input2"];
output = binaryTree(test_set,inputs,"target",5)

###ROC() - Receiving-Operating Characteristics Curve

##Unsupervised Learning - rudyKmeans.js ###kmeans() -