conftrace_

2007 NeurIPS NeurIPS 2007

Boosting the Area under the ROC Curve

Abstract

We show that any weak ranker that can achieve an area under the ROC curve slightly better than 1/2 (which can be achieved by random guessing) can be efﬁ- ciently boosted to achieve an area under the ROC curve arbitrarily close to 1. We further show that this boosting can be performed even in the presence of indepen- dent misclassiﬁcation noise, given access to a noise-tolerant weak ranker.

📈 Trend Setter - Ranking

🐣 Hot Topic Early Bird - supervised learning

Authors

Phil Long, Rocco Servedio

Topics

Machine Learning > Core Methods > Classification Machine Learning > Learning Types > Weakly Supervised Learning Machine Learning > Optimization & Theory > Learning Theory Machine Learning > Optimization & Theory > Optimization Machine Learning > Learning Types > Ensemble Learning Machine Learning > Core Methods > Ranking Deep Learning > Learning Types > Classification

Keywords

binary classification ensemble learning supervised learning boosting roc curve area under roc weak ranker area under curve weak learner area under the roc curve area under the curve misclassification noise

More from NeurIPS 2007

All NeurIPS 2007 papers →

Learning to classify complex patterns using a VLSI network of spiking neurons

Efficient Convex Relaxation for Transductive Support Vector Machine

Message Passing for Max-weight Independent Set

Sparse Feature Learning for Deep Belief Networks

Receding Horizon Differential Dynamic Programming

A Risk Minimization Principle for a Class of Parzen Estimators