caa a22 4500 605478392 CHVBK 20210128100405.0 cr unu---uuuuu 210128e20150901xx s 000 0 eng 10.1007/s10994-014-5478-4 doi (NATIONALLICENCE)springer-10.1007/s10994-014-5478-4 Generalization bounds for learning with linear, polygonal, quadratic and conic side knowledge [Elektronische Daten] [Theja Tulabandhula, Cynthia Rudin] In this paper, we consider a supervised learning setting where side knowledge is provided about the labels of unlabeled examples. The side knowledge has the effect of reducing the hypothesis space, leading to tighter generalization bounds, and thus possibly better generalization. We consider several types of side knowledge, the first leading to linear and polygonal constraints on the hypothesis space, the second leading to quadratic constraints, and the last leading to conic constraints. We show how different types of domain knowledge can lead directly to these kinds of side knowledge. We prove bounds on complexity measures of the hypothesis space for quadratic and conic side knowledge, and show that these bounds are tight in a specific sense for the quadratic case. The Author(s), 2014 Statistical learning theory nationallicence Generalization bounds nationallicence Rademacher complexity nationallicence Covering numbers, constrained linear function classes nationallicence Side knowledge nationallicence Tulabandhula Theja Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 02139, Cambridge, MA, USA aut Rudin Cynthia MIT Sloan School of Management, Massachusetts Institute of Technology, 02139, Cambridge, MA, USA aut Machine Learning Springer US; http://www.springer-ny.com 100/2-3(2015-09-01), 183-216 0885-6125 100:2-3<183 2015 100 10994 https://doi.org/10.1007/s10994-014-5478-4 text/html Onlinezugriff via DOI BK010053 XK010053 XK010000 Metadata rights reserved Springer special CC-BY-NC licence nationallicence 1 research-article jats NATIONALLICENCE NATIONALLICENCE NL-springer NATIONALLICENCE 856 40 https://doi.org/10.1007/s10994-014-5478-4 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Tulabandhula Theja Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 02139, Cambridge, MA, USA aut NATIONALLICENCE 700 1- Rudin Cynthia MIT Sloan School of Management, Massachusetts Institute of Technology, 02139, Cambridge, MA, USA aut NATIONALLICENCE 773 0- Machine Learning Springer US; http://www.springer-ny.com 100/2-3(2015-09-01), 183-216 0885-6125 100:2-3<183 2015 100 10994