VC dimension
Dimensión VC
- Date
- 1971 · confidence medium
- Place
- Moscow, Russia
- Credited to
- Alexey Chervonenkis y Vladimir Vapnik
- Field
- Knowledge & symbols
What was discovered
A measure of the capacity or complexity of a family of classification functions, defined as the largest number of points that family can shatter in every possible way, introduced by Vladimir Vapnik and Alexey Chervonenkis.
What it led to
- foundation of statistical learning theory (VC theory)
- theoretical basis of support vector machines (SVM)
- criterion for assessing overfitting risk in machine-learning models