Lecture 4: Towards Characterizing Learnability
نویسندگان
چکیده
In this lecture, we introduce the notion of the growth function corresponding to a hypothesis class. This is the key notion that allows us to prove the learnability of infinite hypothesis classes.
منابع مشابه
Advanced Course in Machine Learning
The problem of characterizing learnability is the most basic question of learning theory. A fundamental and long-standing answer, formally proven for supervised classification and regression, is that learnability is equivalent to uniform convergence, and that if a problem is learnable, it is learnable via empirical risk minimization. Furthermore, for the problem of binary classification, unifor...
متن کاملSeptember 1989 LIDS - P - 1910 On Metric Entropy , Vapnik - Chervonenkis Dimension , and Learnability for a Class of Distributions 1
In [23], Valiant proposed a formal framework for distribution-free concept learning which has generated a great deal of interest. A fundamental result regarding this framework was proved by Blumer et al. [6] characterizing those concept classes which are learnable in terms of their Vapnik-Chervonenkis (VC) dimension. More recently, Benedek and Itai [4] studied learnability with respect to a fix...
متن کاملLearnability : Basic Definitions , Examples , Results
The notion of learnability plays a central role in learning theory and continues to be an active topic of study. As noted in the last lecture, it is related to the notion of statistical consistency, with origins in theoretical computer science. Over the next few lectures, we will study various models of learnability related to the probably approximately correct (PAC) learning model introduced i...
متن کاملOn Learnability, Complexity and Stability
We consider the fundamental question of learnability of a hypotheses class in the supervised learning setting and in the general learning setting introduced by Vladimir Vapnik. We survey classic results characterizing learnability in term of suitable notions of complexity, as well as more recent results that establish the connection between learnability and stability of a learning algorithm.
متن کاملCS5236 – Advanced Automata Theory
Advanced Automata Theory is a lecture which will first review the basics of formal languages and automata theory and then give insight into specific topics from the theory of automata theory. In computer science, automata are an important tool for many theoretical results and various types of automata have been used to characterise complexity classes. The lecture will give a deeper understandin...
متن کامل