A Boosting Method with Asymmetric Mislabeling Probabilities That Depend on Covariates

• Activity Number: 356
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Learning and Data Mining
• Author(s): Kenichi Hayashi*+
• Companies: Osaka University
• Keywords: Boosting ; Classification ; Asymmetric mislabeling mechanism ; Robustness

Abstract

In this paper, we provide a boosting method for a kind of noisy data where the probability of mislabeling depends on the label of a case. The mechanism of the model is based on a simple idea and gives natural interpretation as a mislabel model. The boosting algorithm is derived from an extension of the exponential loss function, which gives the AdaBoost algorithm. The connection between the proposed method and an asymmetric mislabel model is shown. We show that the loss function proposed constructs a classifier which attains the minimum error rate for a true label. Numerical experiments illustrate how well the proposed method performs in comparison to existing methods.