References & Citations
Computer Science > Machine Learning
Title: On the Robustness of Decision Tree Learning under Label Noise
(Submitted on 20 May 2016 (this version), latest version 26 Aug 2016 (v2))
Abstract: In most practical problems of classifier learning, the training data suffers from the label noise. Hence, it is important to understand how robust is a learning algorithm to such label noise. Experimentally, Decision trees have been found to be more robust against label noise than SVM and logistic regression. This paper presents some theoretical results to show that decision tree algorithms are robust to symmetric label noise under the assumption of large sample size. We also present some sample complexity results for this robustness. Through extensive simulations we illustrate this robustness.
Submission history
From: Naresh Manwani [view email][v1] Fri, 20 May 2016 11:31:26 GMT (32kb)
[v2] Fri, 26 Aug 2016 08:58:06 GMT (58kb)
Link back to: arXiv, form interface, contact.