TY - CONF ID - moewes_tackling_2008 T1 - Tackling Multiple-Instance Problems in Safety-Related Domains by Quasilinear {SVM} A1 - Moewes, Christian A1 - Otte, Clemens A1 - Kruse, Rudolf ED - Dubois, Didier ED - Lubiano, M. Asunción ED - Prade, Henri ED - Gil, María Ángeles ED - Grzegorzewski, Przemysław ED - Hryniewicz, Olgierd TI - Soft Methods for Handling Variability and Imprecision T3 - Advances in Soft Computing Y1 - 2008 VL - 48 SP - 409 EP - 416 PB - Springer Berlin/Heidelberg SN - 978-3-540-85026-7 SN - 1615-3871 UR - http://springerlink.com/content/870h7h003w457740/ M2 - doi: 10.1007/978-3-540-85027-4_49 KW - Multiple-Instance Learning KW - Safety-Related Systems KW - Support Vector Machine N2 - In this paper we introduce a preprocessing method for safety-related applications. Since we concentrate on scenarios with highly unbalanced misclassification costs, we briefly discuss a variation of multiple-instance learning (MIL) and recall soft margin hyperplane classifiers; in particular the principle of a support vector machine (SVM). According to this classifier, we present a training set selection method for learning quasilinear SVMs which guarantee both high accuracy and interpretability to a higher degree. We conclude with annotating on a real-world application and potential extensions for future research in this domain. ER -