Page 1. The Adviceptron: Giving Advice To The Perceptron Gautam Kunapuli1, Kristin P. Bennett2, Richard Maclin3 and Jude W. Shavlik1 Richard Maclin3 and Jude W. Shavlik1 1University of Wisconsin-Madison, USA 2Rensselaer Polytechnic... more
Page 1. The Adviceptron: Giving Advice To The Perceptron Gautam Kunapuli1, Kristin P. Bennett2, Richard Maclin3 and Jude W. Shavlik1 Richard Maclin3 and Jude W. Shavlik1 1University of Wisconsin-Madison, USA 2Rensselaer Polytechnic Institute, USA ...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Research Interests:
... Sriraam Natarajan natarasr@biostat.wisc.edu Prasad Tadepalli* tadepall@cs.orst.edu Gautam Kunapuli kunapg@biostat.wisc.edu Jude Shavlik shavlik@cs.wisc.edu ... 3. Results and Analysis Algorithm UW Citeseer CR-Learner 0.74 0.70 MLN-2... more
... Sriraam Natarajan natarasr@biostat.wisc.edu Prasad Tadepalli* tadepall@cs.orst.edu Gautam Kunapuli kunapg@biostat.wisc.edu Jude Shavlik shavlik@cs.wisc.edu ... 3. Results and Analysis Algorithm UW Citeseer CR-Learner 0.74 0.70 MLN-2 0.5 0.34 MLN-N 0.52 0.40 ...
Research Interests:
Research Interests:
Research Interests:
Research Interests:
Page 1. The Adviceptron: Giving Advice To The Perceptron Gautam Kunapuli1, Kristin P. Bennett2, Richard Maclin3 and Jude W. Shavlik1 Richard Maclin3 and Jude W. Shavlik1 1University of Wisconsin-Madison, USA 2Rensselaer Polytechnic... more
Page 1. The Adviceptron: Giving Advice To The Perceptron Gautam Kunapuli1, Kristin P. Bennett2, Richard Maclin3 and Jude W. Shavlik1 Richard Maclin3 and Jude W. Shavlik1 1University of Wisconsin-Madison, USA 2Rensselaer Polytechnic Institute, USA ...
Research Interests:
We propose a novel approach for incorporating prior knowledge into the online binary support vector classifica- tion problem. An existing advice-taking approach, when prior knowledge is in the form of polyhedral knowledge sets in input... more
We propose a novel approach for incorporating prior knowledge into the online binary support vector classifica- tion problem. An existing advice-taking approach, when prior knowledge is in the form of polyhedral knowledge sets in input space of data, is via knowledge-based support vector machines (KB- SVMs). We adopt the formalism of passive-aggressive algo- rithms to derive an online version of KBSVMs when the advice is fixed for every learning round. The goal is to successively update the decision function taking into account prior knowl- edge in the form of soft polyhedral advice so as to make increas- ingly accurate predictions on subsequent rounds. The advice helps speed up and bias learning so that better generalization can be obtained with less data.
Research Interests:
Research Interests:
Research Interests:
... In Proceedings of the 7th European Conference on Computer Vision-Part IV, ECCV '02, pages 776–792, London, UK, UK, 2002. Springer-Verlag. [18] Kiri Wagstaff, Claire Cardie, Seth Rogers, and Stefan Schroedl. Constrained k-means... more
... In Proceedings of the 7th European Conference on Computer Vision-Part IV, ECCV '02, pages 776–792, London, UK, UK, 2002. Springer-Verlag. [18] Kiri Wagstaff, Claire Cardie, Seth Rogers, and Stefan Schroedl. Constrained k-means clustering with background knowledge. ...