[图书][B] Information geometry and its applications
S Amari - 2016 - books.google.com
This is the first comprehensive book on information geometry, written by the founder of the
field. It begins with an elementary introduction to dualistic geometry and proceeds to a wide …
field. It begins with an elementary introduction to dualistic geometry and proceeds to a wide …
[图书][B] Algorithmic learning in a random world
Vladimir Vovk Alexander Gammerman Glenn Shafer Second Edition Page 1 Vladimir Vovk
Alexander Gammerman Glenn Shafer Algorithmic Learning in a Random World Second …
Alexander Gammerman Glenn Shafer Algorithmic Learning in a Random World Second …
Testing exchangeability: Fork-convexity, supermartingales and e-processes
Suppose we observe an infinite series of coin flips X 1, X 2,…, and wish to sequentially test
the null that these binary random variables are exchangeable. Nonnegative …
the null that these binary random variables are exchangeable. Nonnegative …
Context-tree weighting and Bayesian Context Trees: Asymptotic and non-asymptotic justifications
I Kontoyiannis - IEEE Transactions on Information Theory, 2023 - ieeexplore.ieee.org
The Bayesian Context Trees (BCT) framework is a recently introduced, general collection of
statistical and algorithmic tools for modelling, analysis and inference with discrete-valued …
statistical and algorithmic tools for modelling, analysis and inference with discrete-valued …
[PDF][PDF] The MDL principle, penalized likelihoods, and statistical risk
We determine, for both countable and uncountable collections of functions,
informationtheoretic conditions on a penalty pen (f) such that the optimizer ˆf of the penalized …
informationtheoretic conditions on a penalty pen (f) such that the optimizer ˆf of the penalized …
Active citizenship: Participatory patterns of European youth
Purpose: Treating Active Citizenship as a sum of behavioural indicators requires certain
prerequisites that can be difficult to meet in practice (eg structural validity and measurement …
prerequisites that can be difficult to meet in practice (eg structural validity and measurement …
Time Series Analysis with Bayesian Context Trees: Classical Asymptotics and Finite-n Bounds
I Kontoyiannis - 2023 IEEE International Symposium on …, 2023 - ieeexplore.ieee.org
The Bayesian Context Trees (BCT) framework is a recently introduced, general collection of
statistical and algorithmic tools for modelling, analysis and inference with discrete-valued …
statistical and algorithmic tools for modelling, analysis and inference with discrete-valued …
Asymptotically minimax regret by Bayes mixtures for non-exponential families
J Takeuchi, AR Barron - 2013 IEEE Information Theory …, 2013 - ieeexplore.ieee.org
We study the problems of data compression, gambling and prediction of a sequence xn= x 1
x 2... xn from an alphabet X, in terms of regret with respect to various families of probability …
x 2... xn from an alphabet X, in terms of regret with respect to various families of probability …
Information geometry and its applications
S Amari - J. Math. Psychol, 2005 - Springer
Information geometry is a method of exploring the world of information by means of modern
geometry. Theories of information have so far been studied mostly by using algebraic …
geometry. Theories of information have so far been studied mostly by using algebraic …
MDL, penalized likelihood, and statistical risk
AR Barron, C Huang, JQ Li… - 2008 IEEE Information …, 2008 - ieeexplore.ieee.org
We determine, for both countable and uncountable collections of functions, information-
theoretic conditions on a penalty pen (f) such that the optimizer f of the penalized log …
theoretic conditions on a penalty pen (f) such that the optimizer f of the penalized log …