Inference with the Universum
作者： Jason Weston , Ronan Collobert , Fabian Sinz , Léon Bottou , Vladimir Vapnik
关键词:
摘要: In this paper we study a new framework introduced by Vapnik (1998) and (2006) that is an alternative capacity concept to the large margin approach. particular case of binary classification, are given set labeled examples, collection "non-examples" do not belong either class interest. This collection, called Universum, allows one encode prior knowledge representing meaningful concepts in same domain as problem at hand. We describe algorithm leverage Universum maximizing number observed contradictions, show experimentally approach delivers accuracy improvements over using data alone.
mpg.de
core.ac.uk
bethgelab.org
acm.org参考文章(13)
Vladimir Naumovich Vapnik, Estimation of Dependences Based on Empirical Data ,(2010)
Henry S. Baird, Document image defect models Document image analysis. pp. 315- 325 ,(1995) , 10.1007/978-3-642-77281-8_26
Yves Grandvalet, Stéphane Canu, Stéphane Boucheron, Noise injection: theoretical prospects Neural Computation. ,vol. 9, pp. 1093- 1108 ,(1997) , 10.1162/NECO.1997.9.5.1093
Todd K. Leen, From Data Distributions to Regularization in Invariant Learning neural information processing systems. ,vol. 7, pp. 223- 230 ,(1994) , 10.1162/NECO.1995.7.5.974
Ping Zhong, Masao Fukushima, A new multi-class support vector algorithm Optimization Methods & Software. ,vol. 21, pp. 359- 372 ,(2006) , 10.1080/10556780500094812
O. L. Mangasarian, Linear and Nonlinear Separation of Patterns by Linear Programming Operations Research. ,vol. 13, pp. 444- 452 ,(1965) , 10.1287/OPRE.13.3.444
Bernhard E. Boser, Isabelle M. Guyon, Vladimir N. Vapnik, A training algorithm for optimal margin classifiers conference on learning theory. pp. 144- 152 ,(1992) , 10.1145/130385.130401
Bernhard Schölkopf, Chris Burges, Vladimir Vapnik, Incorporating Invariances in Support Vector Learning Machines international conference on artificial neural networks. pp. 47- 52 ,(1996) , 10.1007/3-540-61510-5_12
J. Shawe-Taylor, P.L. Bartlett, R.C. Williamson, M. Anthony, Structural risk minimization over data-dependent hierarchies IEEE Transactions on Information Theory. ,vol. 44, pp. 1926- 1940 ,(1998) , 10.1109/18.705570
Vladimir Naumovich Vapnik, Vlamimir Vapnik, Statistical learning theory John Wiley & Sons. ,(1998)