Print Email Facebook Twitter Robust domain-adaptive discriminant analysis Title Robust domain-adaptive discriminant analysis Author Kouw, W.M. (TU Delft Pattern Recognition and Bioinformatics; Eindhoven University of Technology) Loog, M. (TU Delft Pattern Recognition and Bioinformatics; University of Copenhagen) Date 2021 Abstract Consider a domain-adaptive supervised learning setting, where a classifier learns from labeled data in a source domain and unlabeled data in a target domain to predict the corresponding target labels. If the classifier’s assumption on the relationship between domains (e.g. covariate shift, common subspace, etc.) is valid, then it will usually outperform a non-adaptive source classifier. If its assumption is invalid, it can perform substantially worse. Validating assumptions on domain relationships is not possible without target labels. We argue that, in order to make domain-adaptive classifiers more practical, it is necessary to focus on robustness; robust in the sense that an adaptive classifier will still perform at least as well as a non-adaptive classifier without having to rely on the validity of strong assumptions. With this objective in mind, we derive a conservative parameter estimation technique, which is transductive in the sense of Vapnik and Chervonenkis, and show for discriminant analysis that the new estimator is guaranteed to achieve a lower risk on the given target samples compared to the source classifier. Experiments on problems with geographical sampling bias indicate that our parameter estimator performs well. Subject Discriminant analysisDomain adaptationRobust estimatorTransduction To reference this document use: http://resolver.tudelft.nl/uuid:b6baa107-2f88-44fe-8ed5-f0bd52b427c8 DOI https://doi.org/10.1016/j.patrec.2021.05.005 ISSN 0167-8655 Source Pattern Recognition Letters, 148, 107-113 Part of collection Institutional Repository Document type journal article Rights © 2021 W.M. Kouw, M. Loog Files PDF 1_s2.0_S0167865521001732_main.pdf 746.15 KB Close viewer /islandora/object/uuid:b6baa107-2f88-44fe-8ed5-f0bd52b427c8/datastream/OBJ/view