MSVM, Poly.Penalized.Likelihood, Nonparametric.LASSO-variable selector

[Grace Wahba]

    Announcing papers recently available via 
    http://www.stat.wisc.edu/~wahba 
    click on TRLIST

   1,2 and 3 below present the Multicategory Support Vector Machine (MSVM),  
   a generalization of the SVM which classifies to one of k categories via 
   a single optimization problem. 4 contrasts the MSVM with the 
   Polychotomous Penalized Likelihood estimate, which estimates 
   k probabilities, one for each category. 5 and 6  present a new
   nonparametric variable selection and model building via likelihood
   basis pursuit and a generalization of the LASSO.

   1 Lee, Y., Lin, Y. and Wahba, G. " Multicategory Support Vector Machines,
     Theory, and Application to the Classification of Microarray Data and
     Satellite Radiance Data " TR 1064, September 2002.

   2  Lee, Y. " Multicategory Support Vector Machines, Theory, and
     Application to the Classification of Microarray Data and Satellite
     Radiance Data " TR 1063, September 2002. PhD. Thesis.

   3 Lee, Y. and Lee, C.-K. Classification of Multiple Cancer Types by
     Multicategory Support Vector Machines Using Gene Expression Data. (ps)
     TR 1051, April 2002, minor revisions July 2002.

   4 Wahba, G. " Soft and Hard Classification by Reproducing Kernel Hilbert
     Space Methods " (ps) (pdf) TR 1067, October 2002. To appear Proceedings
     of the National Academy of Sciences.

   5 Zhang, H. " Nonparametric Variable Selection and Model Building Via
     Likelihood Basis Pursuit " TR 1066, September 2002. PhD. Thesis.

   6 Zhang, H., Wahba, G., Lin, Y., Voelker, M., Ferris, M., Klein, R. and
     Klein, B. Variable Selection and Model Building via Likelihood Basis
     Pursuit (ps) (pdf) TR 1059, July, 2002.