ComputingComputer ScienceData ScienceFeature EngineeringMachine LearningStatistical LearningClassificationFeature SelectionKernel MethodsPattern RecognitionRepresentation LearningSupport Vector MachinesKernel SelectionMixture Of ClassifiersMulti-task Feature LearningMulti-task Feature SelectionMultitask LearningSupport Vector ClassificationSupport Vector Machine ClassificationSvms
We compute a common feature selection or kernel selection configuration for multiple support vector machines (SVMs) trained on different yet inter-related datasets. The method is advantageous when multiple classification tasks and differently labeled datasets exist over a common input space. Different datasets can mutually reinforce a common choice of representation or relevant features for their various classifiers. We derive a multi-task representation learning approach using the maximum entropy discrimination formalism. The resulting convex algorithms maintain the global solution properties of support vector machines. However, in addition to multiple SVM classification/regression parameters they also jointly estimate an optimal subset of features or optimal combination of kernels. Experiments are shown on standardized datasets.
7
Regularized multi--task learning
Theodoros Evgeniou, Massimiliano Pontil · 2004 · 1.5K citations
Jason Weston, Sayan Mukherjee, Olivier Chapelle et al. · 2000 · 979 citations
A Model of Inductive Bias Learning
2000 · 781 citations
Learning the Kernel Matrix with Semi-Definite Programming
Gert Lanckriet, Peter L. Bartlett, Laurent El Ghaoui et al. · Explore Bristol Research · 2002 · 388 citations · Full text
Model Selection for Support Vector Machines
Olivier Chapelle, Vladimir Vapnik · 1999 · 349 citations