Applying MDL to learn best model granularity

DOI10.1016/S0004-3702(00)00034-5zbMath0948.68092OpenAlexW2170052858WikidataQ128012734 ScholiaQ128012734MaRDI QIDQ1583224

Q. Gao, Paul M. B. Vitányi, Ming Li

Publication date: 26 October 2000

Published in: Artificial Intelligence (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/s0004-3702(00)00034-5

zbMATH Keywords

Kolmogorov complexityfeedforward neural networkOccam's razorBayes' ruleuniversal priorimproved elastic matchinglearning best model granularitylearning optimal feature extraction intervallearning optimal hidden layerMinimum Description Length principle (MDL)modeling robot armon-line handwritten character recognition

Mathematics Subject Classification ID

Algorithmic information theory (Kolmogorov complexity, etc.) (68Q30)

Related Items (2)

Kolmogorov's Last Discovery? (Kolmogorov and Algorithmic Statistics) ⋮ The generalized universal law of generalization.

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Inferring decision trees using the minimum description length principle
• Modeling by shortest data description
• Inductive reasoning and Kolmogorov complexity
• The miraculous universal distribution
• Deductive learning
• Universal coding, information, prediction, and estimation
• A theory of the learnable
• Complexity-based induction systems: Comparisons and convergence theorems
• Minimum description length induction, Bayesianism, and Kolmogorov complexity
• The minimum description length principle in coding and modeling
• An Information Measure for Classification
• The definition of random sequences
• A formal theory of inductive inference. Part I

This page was built for publication: Applying MDL to learn best model granularity