History
In 1995 Corinna Cortes and Vladimir Vapnik published their work on a supervised learning algorithm called:
Support Vector Machines
Card 1 of 259. Answer: Support Vector Machines
Support Vector Machines
Card 1 of 259. Answer: Support Vector Machines
About · Quizzes · History of AI · To suggest new questions or report an error send me a dm.
Support vector machines. Cortes and Vapnik's paper "Support-Vector Networks" appeared in Machine Learning in 1995. An SVM is a classifier that finds the hyperplane separating two classes with the widest margin, the distance to the nearest training points, which are called the support vectors.
The idea was not entirely new. Vapnik and colleagues had described maximum-margin linear classifiers in the 1960s, and Boser, Guyon and Vapnik added the kernel trick in 1992. Cortes and Vapnik's main addition was the soft margin, for data that cannot be separated perfectly: minimize ½‖w‖² + C Σ ξᵢ subject to yᵢ(w·xᵢ + b) ≥ 1 − ξᵢ and ξᵢ ≥ 0. The margin width is 2/‖w‖.
Interviewers ask what C does: a large C punishes violations and gives a narrow margin that can overfit; a small C allows violations for a wider margin. Expect questions on kernels such as the RBF, K(x, z) = exp(−γ‖x − z‖²), on feature scaling, and on why kernel SVMs struggle beyond tens of thousands of rows.