Sciweavers

Free Online Productivity Tools i2Speak i2Symbol i2OCR iTex2Img iWeb2Print iWeb2Shot i2Type iPdf2Split iPdf2Merge i2Bopomofo i2Arabic i2Style i2Image i2PDF iLatex2Rtf Sci2ools

9

ICML
2005
IEEE

favoriteEmaildiscussreport

99views Machine Learning» more ICML 2005»

Predicting good probabilities with supervised learning

14 years 5 months ago

Predicting good probabilities with supervised learning

Download www.machinelearning.org

We examine the relationship between the predictions made by different learning algorithms and true posterior probabilities. We show that maximum margin methods such as boosted trees and boosted stumps push probability mass away from 0 and 1 yielding a characteristic sigmoid shaped distortion in the predicted probabilities. Models such as Naive Bayes, which make unrealistic independence assumptions, push probabilities

Alexandru Niculescu-Mizil, Rich Caruana

Real-time Traffic

ICML 2005 | Machine Learning | Stumps Push Probability | True Posterior Probabilities | Unrealistic Independence Assumptions |

claim paper

Related Content

» A Variational EM Approach to Predicting Uncertainty in Supervised Learning

» Supervised Aggregation of Classifiers using Artificial Prediction Markets

» SelfSupervised Chinese Word Segmentation

» Bayesian Methods for Support Vector Machines Evidence and Predictive Class Probabilities

» Supervised Learning of a Probabilistic Lexicon of Verb Semantic Classes

» Structured Output Learning with Indirect Supervision

» Bounds for Multistage Stochastic Programs Using Supervised Learning Strategies

» Using Clustering to Learn Distance Functions for Supervised Similarity Assessment

» Multivariate Discretization by Recursive Supervised Bipartition of Graph

Post Info
More Details (n/a)

Added	17 Nov 2009
Updated	17 Nov 2009
Type	Conference
Year	2005
Where	ICML
Authors	Alexandru Niculescu-Mizil, Rich Caruana

Comments (0)