Support vector machines

• Guenther, Nick ^[1] ; Schonlau, Matthias ^[1]
  1. [1] University of Waterloo

     University of Waterloo

     Canadá

     
• Localización: The Stata journal, ISSN 1536-867X, Vol. 16, Nº. 4, 2016, págs. 917-937
• Idioma: inglés
• DOI: 10.1177/1536867x1601600407
• Enlaces
  • Texto completo
• Resumen

  • Support vector machines are statistical- and machine-learning techniques with the primary goal of prediction. They can be applied to continuous, binary, and categorical outcomes analogous to Gaussian, logistic, and multinomial regression. We introduce a new command for this purpose, svmachines. This package is a thin wrapper for the widely deployed libsvm (Chang and Lin, 2011, ACM Transactions on Intelligent Systems and Technology 2(3): Article 27). We illustrate svmachines with two examples.