Yes (223)
Used filters
(3)
Number of found documents:
223
Published from
to
Fast Dependency-Aware Feature Selection in Very-High-Dimensional Pattern Recognition Problems
Somol, Petr; Grim, Jiří
2011 - English
The paper addresses the problem of making dependency-aware feature selection feasible in pattern recognition problems of very high dimensionality. The idea of individually best ranking is generalized to evaluate the contextual quality of each feature in a series of randomly generated feature subsets. Each random subset is evaluated by a criterion function of arbitrary choice (permitting functions of high complexity). Eventually, the novel dependency-aware feature rank is computed, expressing the average benefit of including a feature into feature subsets. The method is efficient and generalizes well especially in very-high-dimensional problems, where traditional context-aware feature selection methods fail due to prohibitive computational complexity or to over-fitting. The method is shown well capable of over-performing the commonly applied individual ranking which ignores important contextual information contained in data.
Keywords:
feature selection,; high dimensionality; ranking; generalization; over-fitting; stability; classification; pattern recognition; machine learning
Fulltext is available at external website.
Fast Dependency-Aware Feature Selection in Very-High-Dimensional Pattern Recognition Problems
The paper addresses the problem of making dependency-aware feature selection feasible in pattern recognition problems of very high dimensionality. The idea of individually best ranking is generalized ...
Bayesian Methods for Optimization of Radiation Monitoring Networks
Šmídl, Václav; Hofman, Radek
2011 - English
Release of radioactive material into the atmosphere is the last possible resort of any accident in a nuclear power plant. It is an extremely rare event, however with severe consequences for potentially many people living in proximity of the power plant. Awareness of radiation security has been increased after the Chernobyl accident, and almost every country is now equipped with monitoring network of on-line connected receptors continually measuring radiation levels. Initial configurations of the network were designed by experts using their experience.In this report, we are concerned with local scale modeling of less severe accident in the range of tens of kilometers from the nuclear power plant. Both the stationary and mobile groups will be discussed. The preferred model of uncertainty is the empirical density which will be assimilated with measurements using the sequential Monte Carlo methodology. We will discuss influence of various loss functions.
Keywords:
radiation monitoring; UAV; data assimilation
Fulltext is available at external website.
Bayesian Methods for Optimization of Radiation Monitoring Networks
Release of radioactive material into the atmosphere is the last possible resort of any accident in a nuclear power plant. It is an extremely rare event, however with severe consequences for ...
Monitorování radiace v časné fázi nehody na jaderném zařízení - analýza všech typů měření použitelných pro korekci modelových předpovědí
Pecha, Petr; Kuča, Petr; Češpírová, Irena; Hofman, Radek
2011 - Czech
Zpráva se zaměřuje na metody monitorování radioaktivního znečistění v časné fázi nehody spojené s únikem radioaktivity do životního prostředí. Cílem prováděného rozboru je určit ty metody monitorování prováděné stávajícími radiačními sítěmi, které mohou poskytovat měření z terénu pro jejich další využití v oblasti zlepšování modelových předpovědí vývoje radiační situace. Pro tyto účely byly vyvinuty speciální statistické metody bayesovské filtrace provádějící asimilaci modelových předpovědí s měřeními z terénu.
Keywords:
data assimilation; radiation monitoring; nuclear safety
Fulltext is available at external website.
Monitorování radiace v časné fázi nehody na jaderném zařízení - analýza všech typů měření použitelných pro korekci modelových předpovědí
Zpráva se zaměřuje na metody monitorování radioaktivního znečistění v časné fázi nehody spojené s únikem radioaktivity do životního prostředí. Cílem prováděného rozboru je určit ty metody monitorování ...
On polyhedral approximations of polytopes for learning Bayes nets
Studený, Milan; Haws, D.
2011 - English
We review three vector encodings of Bayesian network structures. The first one has recently been applied by Jaakkola et al., the other two use special integral vectors, called imsets. The central topic is the comparison of outer polyhedral approximations of the corresponding polytopes. We show how to transform the inequalities suggested by Jaakkola et al. to the framework of imsets. The result of our comparison is the observation that the implicit polyhedral approximation of the standard imset polytope suggested in (Studený Vomlel 2010) gives a closer approximation than the (transformed) explicit polyhedral approximation from (Jaakkola et al. 2010). Finally, we confirm a conjecture from (Studený Vomlel 2010) that the above-mentioned implicit polyhedral approximation of the standard imset polytope is an LP relaxation of the polytope.
Keywords:
learning Bayesian networks; imsets; polytopes
Fulltext is available at external website.
On polyhedral approximations of polytopes for learning Bayes nets
We review three vector encodings of Bayesian network structures. The first one has recently been applied by Jaakkola et al., the other two use special integral vectors, called imsets. The central ...
Stable distributions: On parametrizations of characteristic exponent
Karlová, Andrea
2011 - English
In this report we investigate theory of stable distributions and their role in probability theory. We are interested in derivation of canonical measure, semigroup operator and mainly in parametrizations of characteristic exponents. We finally introduce a new parametrization.
Keywords:
stable distribution; characteristic function; characteristic exponent
Fulltext is available at external website.
Stable distributions: On parametrizations of characteristic exponent
In this report we investigate theory of stable distributions and their role in probability theory. We are interested in derivation of canonical measure, semigroup operator and mainly in ...
Approximate Dynamic Programming based on High Dimensional Model Representation
Pištěk, Miroslav
2011 - English
In this article, an efficient algorithm for an optimal decision strategy approximation is introduced. The proposed approximation of the Bellman equation is based on HDMR technique. This non-parametric function approximation is used not only to reduce memory demands necessary to store Bellman function, but also to allow its fast approximate minimization. On that account, a clear connection between HDMR minimization and discrete optimization is newly established. In each time step of the backward evaluation of the Bellman function, we relax the parameterized discrete minimization subproblem to obtain parameterized trust region problem. We observe that the involved matrix is the same for all parameters owning to the structure of HDMR approximation. We find eigenvalue decomposition of this matrix to solve all trust region problems effectively.
Keywords:
HDMR approximation; Bellman equation; minimization of HDMR functions
Fulltext is available at external website.
Approximate Dynamic Programming based on High Dimensional Model Representation
In this article, an efficient algorithm for an optimal decision strategy approximation is introduced. The proposed approximation of the Bellman equation is based on HDMR technique. This non-parametric ...
Notes on projection based modelling of beta-distributed weights of a two-component mixture
Dedecius, Kamil
2011 - English
This report contains brief notes on estimation of beta-distributed weight of a Gaussian mixture. The results are directly applied in paper Kárný, M.: On approximate Bayesian recursive estimation]. First, we develop a method to update the beta distribution of weights by new data (evidences) and show, that a projection is needed to preserve the low modelling complexity. Then, we show how forgetting may be applied to improve adaptivity. The results can be immediately applied to multicomponent mixtures.
Keywords:
beta mixtures; projection; Bayesian modelling
Fulltext is available at external website.
Notes on projection based modelling of beta-distributed weights of a two-component mixture
This report contains brief notes on estimation of beta-distributed weight of a Gaussian mixture. The results are directly applied in paper Kárný, M.: On approximate Bayesian recursive estimation]. ...
Evaluation of tight bounds for divergences
Harremoes, P.; Vajda, Igor
2010 - English
The paper presents a general method for evaluation of the joint range of pairs of f-divergences. This range provides tight maxima and minima for one f-divergence for given value of the other. Applications in information theory, identification and detection are mentioned. Práce prezentuje obecnou metodu pro stanovení oblasti hodnot dvojic f-divergencí. Tato oblast poskytuje těsná maxima a minima jedné divergence pro danou hodnotu druhé. Jsou zmíněny aplikace takových mezí v teorii informace, identifikaci a detekci.
Keywords:
Divergence bounds; Conditional divergence maxima; Conditional divergence minima
Fulltext is available at external website.
Evaluation of tight bounds for divergences
The paper presents a general method for evaluation of the joint range of pairs of f-divergences. This range provides tight maxima and minima for one f-divergence for given value of the other. ...
Vyhodnocení markovských řetězců pomocí funkce MR,M v Matlabu
Michálek, Jiří; Šiman, Miroslav
2010 - Czech
Výzkumná zpráva se zabývá vyhodnocováním vlastností markovských či semimarkovských řetězců pomocí vhodné procedury zkonstruobané v Matlabu The research report deals with the evaluation of Markov or semi-Markov chains using a suitable procedure constructed in Matlab
Keywords:
Markov chain; Semi-Markov chain; Program in Matlab
Fulltext is available at external website.
Vyhodnocení markovských řetězců pomocí funkce MR,M v Matlabu
Výzkumná zpráva se zabývá vyhodnocováním vlastností markovských či semimarkovských řetězců pomocí vhodné procedury zkonstruobané v Matlabu...
Goodness-of-Fit Disparity Statistics Obtained by Hypothetical and Empirical Quantizations
Boček, Pavel; Vajda, Igor; van der Meulen, E.
2010 - English
Goodness-of-fit disparity statistics are defined as appropriately scaled phi-disparities or phi-divergences of quantized hypothetical and empirical distributions. It is shown that the classical Pearson-type statistics are obtained if we quantize by means of hypothetical percentiles, and that new spacings-based disparity statistics are obtained if we quantize by means of empirical percentiles.
Keywords:
power divergences; goodness-of-fit; asymptotic normality,
Fulltext is available at external website.
Goodness-of-Fit Disparity Statistics Obtained by Hypothetical and Empirical Quantizations
Goodness-of-fit disparity statistics are defined as appropriately scaled phi-disparities or phi-divergences of quantized hypothetical and empirical distributions. It is shown that the classical ...
NRGL provides central access to information on grey literature produced in the Czech Republic in the fields of science, research and education. You can find more information about grey literature and NRGL at service web
Send your suggestions and comments to nusl@techlib.cz
Provider
Other bases