EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0538
Title: Robustness of kernel-based pairwise learning Authors:  Patrick Gensler - University of Bayreuth (Germany) [presenting]
Andreas Christmann - University of Bayreuth (Germany)
Abstract: Pairwise learning can be applied in a variety of situations such as ranking, which is an important topic in machine learning and information retrieval, and also similarity learning and distance metric learning. Many results on the statistical robustness of kernel-based pairwise learning can be derived under basically no assumptions on the input and output spaces. In particular, neither moment conditions on the conditional distribution of $Y$ given $X = x$ nor the boundedness of the output space is needed. Results on the existence and boundedness of the influence function have been obtained and show the qualitative robustness of the kernel-based estimator.