A Generalization of Sauer’s Lemma to Classes of Large-Margin Functions
نویسنده
چکیده
We generalize Sauer’s Lemma to finite V C-dimension classes H of binary-valued functions on [n] = {1, . . . , n} which have a margin of at least N on every element in a sample S ⊆ [n] of cardinality l, where the margin μh(x) of h ∈ H on a point x ∈ [n] is defined as the largest non-negative integer a such that h is constant on the interval Ia(x) = [x− a, x+ a].
منابع مشابه
A Constrained version of Sauer’s Lemma
Sauer’s Lemma is extended to classes H of binary-valued functions on [n] = {1, . . . , n} which have a margin less than or equal to N on x ∈ [n], where the margin μh(x) of a binary valued function h at a point x ∈ [n] is defined as the largest nonnegative integer a such that h is constant on the interval Ia(x) = [x−a, x+a] ⊆ [n]. Estimates are obtained for the cardinality of classes of binary v...
متن کاملOn the complexity of constrained VC-classes
Sauer’s Lemma is extended to classes HN of binary-valued functions h on [n] = {1, . . . , n} which have a margin less than or equal to N on all x ∈ [n] with h(x) = 1, where the margin μh(x) of h at x ∈ [n] is defined as the largest non-negative integer a such that h is constant on the interval Ia(x) = [x− a, x+ a] ⊆ [n]. Estimates are obtained for the cardinality of classes of binary valued fun...
متن کاملVC Theory of Large Margin Multi-Category Classifiers
In the context of discriminant analysis, Vapnik’s statistical learning theory has mainly been developed in three directions: the computation of dichotomies with binary-valued functions, the computation of dichotomies with real-valued functions, and the computation of polytomies with functions taking their values in finite sets, typically the set of categories itself. The case of classes of vect...
متن کاملConstrained versions of Sauer's lemma
Let [n] = {1, . . . , n}. For a function h : [n] → {0, 1}, x ∈ [n] and y ∈ {0, 1} define by the width ωh(x, y) of h at x the largest non-negative integer a such that h(z) = y on x− a ≤ z ≤ x+ a. We consider finite VC-dimension classes of functions h constrained to have a width ωh(xi, yi) which is larger than N for all points in a sample ζ = {(xi, yi)}1 or a width no larger than N over the whole...
متن کاملOn the VC-dimension and Boolean functions with long runs
The Vapnik-Chervonenkis (VC) dimension and the Sauer-Shelah lemma have found applications in numerous areas including set theory, combinatorial geometry, graph theory and statistical learning theory. Estimation of the complexity of discrete structures associated with the search space of algorithms often amounts to estimating the cardinality of a simpler class which is effectively induced by som...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2004