Sample size lower bounds in PAC learning by algorithmic complexity theory

This paper focuses on a general setup for obtaining sample size lower bounds for learning concept classes under fixed distribution laws in an extended PAC learning framework. These bounds do not depend on the running time of learning procedures and are information-theoretic in nature. They are based...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Theoretical computer science Jg. 209; H. 1; S. 141 - 162
Hauptverfasser: Apolloni, B., Gentile, C.
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Amsterdam Elsevier B.V 06.12.1998
Elsevier
Schlagworte:
Algorithmics. Computability. Computer arithmetics
Applied sciences
Artificial intelligence
Computational learning
Computer science; control theory; systems
Exact sciences and technology
Kolmogorov complexity
Learning and adaptive systems
Sample complexity
Theoretical computing
Computational learning
Kolmogorov complexity
Sample complexity
Markov chain
Lower bound
Kolmogorov law
Taylor series
Conditional probability
Learning algorithm
Motor learning
Computational complexity
Artificial intelligence
ISSN:0304-3975, 1879-2294
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper focuses on a general setup for obtaining sample size lower bounds for learning concept classes under fixed distribution laws in an extended PAC learning framework. These bounds do not depend on the running time of learning procedures and are information-theoretic in nature. They are based on incompressibility methods drawn from Kolmogorov Complexity and Algorithmic Probability theories.
ISSN:0304-3975
1879-2294
DOI:10.1016/S0304-3975(97)00102-3