How does predict work for classification tree algorithm?

Following classification with ClassificationTree.fit I get an object obj which contains all model parameters. Then when using predict(data,model) I try it on new data. Question is that I need to translate the predict code into human language, in order to transform it into c code. The code does Not support predict. Can anyone direct me to the full algorithm predict performs? (I couldn't figure it out from predict.m directly..) Many many thanks Hadas

Answer 1

[Shashank Prasanna]

Type:

>> view(obj)

you can visualize the tree.

predict generates predictions by following the branches of tree until it reaches a leaf node or a missing value. If predict reaches a leaf node, it returns the classification of that node.

This information is also available in the documentation, scroll down to Algorithm section:

http://www.mathworks.com/help/stats/compactclassificationtree.predict.html

Also here is some theory that you may find useful:

http://www.mathworks.com/help/stats/classification-trees-and-regression-trees.html#bsw6a62

Answer 2

[Hadas Lapid]

view(tree) was exactly what I needed, thank you!