13-PracticalMachineLearning

Decision trees vs svm hase et althe elements of

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: friend yes got electricity? read book got new dvd? play computer watch tv Decision Trees •  •  •  •  Fast training Fast prediciton Easy to understand Easy to interpret Decision Tree - Idea E B C A A B C D E Bishop, “Pa4ern Recogni%on and Machine Learning”, Springer, 2006 D Decision Tree - Predic%on E B C A A B C D E D Decision Tree - Training •  Learn the tree structure: –  which feature to query –  which threshold to choose A B C D E Node Purity 10 7 E 7 2 3 5 B 3 2 C A A B C D E D When to Stop •  •  •  •  •  node contains only one class node contains less than x data points max depth is reached node purity is sufficient you start to overfit => cross- valida%on Decision Trees - Disadvantages •  Sensi%ve to small changes in the data •  Overfitng •  Only axis aligned splits Decision Trees vs SVM Has%e et al.,”The Elemen...
View Full Document

This document was uploaded on 12/22/2013.

Ask a homework question - tutors are online