Results 161 to 170 of about 14,721 (214)
Some of the next articles are maybe not open access.
2010
We show that within the Gold paradigm for language learning an informer for a superfinite set can cause an optimal MDL learner to make an infinite amount of mind changes. In this setting an optimal learner can make an infinite amount of wrong choices without approximating the right solution.
Adriaans, P., Mulder, W.
openaire +3 more sources
We show that within the Gold paradigm for language learning an informer for a superfinite set can cause an optimal MDL learner to make an infinite amount of mind changes. In this setting an optimal learner can make an infinite amount of wrong choices without approximating the right solution.
Adriaans, P., Mulder, W.
openaire +3 more sources
2007 IEEE International Symposium on Information Theory, 2007
We point out a potential weakness in the application of the celebrated minimum description length (MDL) principle for model selection. Specifically, it is shown that (although the index of the model class which actually minimizes a two-part code has many desirable properties) a model which has a shorter two- part code-length than another is not ...
P. Adriaans, P.M.B. Vitányi (Paul)
openaire +5 more sources
We point out a potential weakness in the application of the celebrated minimum description length (MDL) principle for model selection. Specifically, it is shown that (although the index of the model class which actually minimizes a two-part code has many desirable properties) a model which has a shorter two- part code-length than another is not ...
P. Adriaans, P.M.B. Vitányi (Paul)
openaire +5 more sources
Proceedings 15th International Conference on Pattern Recognition. ICPR-2000, 2002
We propose an approach to constructing multiple eigenspaces from a set of training images based on the minimum description length (MDL) principle. The main idea is to systematically build a redundant set of eigenspaces, which are treated as hypotheses that are then subject to a selection procedure.
Ales Leonardis, Horst Bischof
openaire +2 more sources
We propose an approach to constructing multiple eigenspaces from a set of training images based on the minimum description length (MDL) principle. The main idea is to systematically build a redundant set of eigenspaces, which are treated as hypotheses that are then subject to a selection procedure.
Ales Leonardis, Horst Bischof
openaire +2 more sources
IEEE Transactions on Information Theory, 2000
Summary: The so-called denoising problem, relative to normal models for noise, is formalized such that ``noise'' is defined as the incompressible part in the data while the compressible part defines the meaningful information-bearing signal. Such a decomposition is effected by minimization of the ideal code length, called for by the minimum description
openaire +1 more source
Summary: The so-called denoising problem, relative to normal models for noise, is formalized such that ``noise'' is defined as the incompressible part in the data while the compressible part defines the meaningful information-bearing signal. Such a decomposition is effected by minimization of the ideal code length, called for by the minimum description
openaire +1 more source
Refinements of MDL and MML Coding
The Computer Journal, 1999Summary: We discuss Rissanen's scheme of `complete coding' in which a two-part data code is further shortened by conditioning the second part not only on the estimates, but also on the fact that these estimates were preferred to any others. We show that the scheme does not lead to improved estimates of parameters.
Chris S. Wallace, David L. Dowe
openaire +1 more source
2003
We point out that a number of standard sample complexity bounds (VC-dimension, PAC-Bayes, and others) are all related to the number of bits required to communicate the labels given the unlabeled data for a natural communication game. Motivated by this observation, we give a general sample complexity bound based on this game that allows us to unify ...
Avrim Blum, John Langford 0001
openaire +1 more source
We point out that a number of standard sample complexity bounds (VC-dimension, PAC-Bayes, and others) are all related to the number of bits required to communicate the labels given the unlabeled data for a natural communication game. Motivated by this observation, we give a general sample complexity bound based on this game that allows us to unify ...
Avrim Blum, John Langford 0001
openaire +1 more source
Using MDL for Grammar Induction
2006In this paper we study the application of the Minimum Description Length principle (or two-part-code optimization) to grammar induction in the light of recent developments in Kolmogorov complexity theory. We focus on issues that are important for construction of effective compression algorithms.
Adriaans, P.W., Jacobs, C.J.H.
openaire +4 more sources
SSRN Electronic Journal, 2020
Over the past 50 years, multidistrict litigation (MDL) has quietly revolutionized civil procedure. MDLs include the largest tort cases in U.S. history, but without the authority of the class-action rule, MDL judges—who formally have only pretrial jurisdiction over individual cases—have resorted to extraordinary procedural exceptionalism to settle ...
Burch, Elizabeth Chamblee, Gluck, Abbe
openaire +1 more source
Over the past 50 years, multidistrict litigation (MDL) has quietly revolutionized civil procedure. MDLs include the largest tort cases in U.S. history, but without the authority of the class-action rule, MDL judges—who formally have only pretrial jurisdiction over individual cases—have resorted to extraordinary procedural exceptionalism to settle ...
Burch, Elizabeth Chamblee, Gluck, Abbe
openaire +1 more source