Auer, Peter; Ortner, Ronald A boosting approach to multiple instance learning. (English) Zbl 1132.68525 Boulicaut, J.-F. (ed.) et al., Machine learning: ECML 2004. 15th European conference on machine learning, Pisa, Italy, September 20–24, 2004, Proceedings. Berlin: Springer (ISBN 978-3-540-23105-9/pbk). Lecture Notes in Computer Science 3201. Lecture Notes in Artificial Intelligence, 63-74 (2004). Summary: In this paper we present a boosting approach to multiple instance learning. As weak hypotheses we use balls (with respect to various metrics) centered at instances of positive bags. For the \(\infty \)-norm these hypotheses can be modified into hyper-rectangles by a greedy algorithm. Our approach includes a stopping criterion for the algorithm based on estimates for the generalization error. These estimates can also be used to choose a preferable metric and data normalization. Compared to other approaches our algorithm delivers improved or at least competitive results on several multiple instance benchmark data sets.For the entire collection see [Zbl 1131.68005]. Cited in 5 Documents MSC: 68T05 Learning and adaptive systems in artificial intelligence PDFBibTeX XMLCite \textit{P. Auer} and \textit{R. Ortner}, Lect. Notes Comput. Sci. 3201, 63--74 (2004; Zbl 1132.68525) Full Text: DOI