Ne (361)
Ano (179)
Použitá omezení
(3)
Počet nalezených dokumentů:
540
Publikováno od
do
Feature Selection - A Very Compact Survey Over the Diversity of Existing Approaches
Somol, Petr; Novovičová, Jana; Pudil, Pavel; Kittler, J.
2010 - anglický
Feature Selection has been a subject of extensive research that nowadays extends far beyond the boundaries of statistical pattern recognition. We provide a concise yet wide view of the topic including representative references in an attempt to point out that important results can be easily overlooked or duplicated in a variety of – even indirectly related – research fields.
Klíčová slova:
feature selection; dimensionality reduction; survey
Plné texty jsou dostupné na jednotlivých ústavech Akademie věd ČR.
Feature Selection - A Very Compact Survey Over the Diversity of Existing Approaches
Feature Selection has been a subject of extensive research that nowadays extends far beyond the boundaries of statistical pattern recognition. We provide a concise yet wide view of the topic ...
On weak solutions of stochastic differential equations
Hofmanová, M.; Seidler, Jan
2010 - anglický
A new proof of existence of weak solutions to stochastic differential equations with continuous coefficients based on ideas from infinite-dimensional stochastic analysis is presented.
Klíčová slova:
weak solutions; stochastic differential equations
Dokument je dostupný na externích webových stránkách.
On weak solutions of stochastic differential equations
A new proof of existence of weak solutions to stochastic differential equations with continuous coefficients based on ideas from infinite-dimensional stochastic analysis is presented.
Implementation of partial forgetting in Mixtools
Dedecius, Kamil
2010 - anglický
The report describes the implementation of the partial forgetting in the Mixtools software package.
Klíčová slova:
partial forgetting
Dokument je dostupný na externích webových stránkách.
Implementation of partial forgetting in Mixtools
The report describes the implementation of the partial forgetting in the Mixtools software package.
Example creation system
Suzdaleva, Evgenia; Nagy, Ivan; Dungl, Martin
2010 - anglický
The technical report describes a system for creation of examples for fully probabilistic dynamic decision making. It represents a documentation for author who wishes to create examples for PhD students or to add their new developed algorithms to the system.
Klíčová slova:
dynamic decision making; example creation; matlab
Dokument je dostupný na externích webových stránkách.
Example creation system
The technical report describes a system for creation of examples for fully probabilistic dynamic decision making. It represents a documentation for author who wishes to create examples for PhD ...
Automatic Removal of Sparse Artifacts in Electroencephalogram
Zima, Miroslav; Tichavský, Petr; Krajča, V.
2010 - anglický
This report presents an algorithm for removing artifacts from EEG signal, which is based on the method of independent component analysis utilizing the signal nonstationarity or sparsity of the artifacts. The algorithm is computationally very fast, enables online processing of long data records with excellent separation accuracy. The algorithm also incorporates using wavelet denoising of the artifact components, recently proposed by Castellanos and Makarov, which reduces distortion of the cleaned data.
Klíčová slova:
electroencephalogram; independent component analysis; artifact removal; wavelet denoising
Dokument je dostupný na externích webových stránkách.
Automatic Removal of Sparse Artifacts in Electroencephalogram
This report presents an algorithm for removing artifacts from EEG signal, which is based on the method of independent component analysis utilizing the signal nonstationarity or sparsity of the ...
Introduction to Feature Selection Toolbox 3 – The C++ Library for Subset Search, Data Modeling and Classification
Somol, Petr; Vácha, Pavel; Mikeš, Stanislav; Hora, Jan; Pudil, Pavel; Žid, Pavel
2010 - anglický
We introduce a new standalone widely applicable software library for feature selection (also known as attribute or variable selection), capable of reducing problem dimensionality to maximize the accuracy of data models, performance of automatic decision rules as well as to reduce data acquisition cost. The library can be exploited by users in research as well as in industry. Less experienced users can experiment with different provided methods and their application to real-life problems, experts can implement their own criteria or search schemes taking advantage of the toolbox framework. In this paper we first provide a concise survey of a variety of existing feature selection approaches. Then we focus on a selected group of methods of good general performance as well as on tools surpassing the limits of existing libraries. We build a feature selection framework around them and design an object-based generic software library. We describe the key design points and properties of the library.
Klíčová slova:
feature selection; software library; subset search; attribute selection; variable selection; optimization; machine learning; classification; pattern recognition
Dokument je dostupný na externích webových stránkách.
Introduction to Feature Selection Toolbox 3 – The C++ Library for Subset Search, Data Modeling and Classification
We introduce a new standalone widely applicable software library for feature selection (also known as attribute or variable selection), capable of reducing problem dimensionality to maximize the ...
How Does Monetary Policy Change? Evidence on Inflation Targeting Countries
Baxa, Jaromír; Horváth, R.; Vašíček, B.
2010 - anglický
We examine the evolution of monetary policy rules in a group of inflation targeting countries (Australia, Canada, New Zealand, Sweden and the United Kingdom), applying a moment-based estimator in a time-varying parameter model with endogenous regressors. Using this novel flexible framework, our main findings are threefold. First, monetary policy rules change gradually, pointing to the importance of applying a time-varying estimation framework. Second, the interest rate smoothing parameter is much lower than typically reported by previous time-invariant estimates of policy rules. External factors matter for all countries, although the importance of the exchange rate diminishes after the adoption of inflation targeting. Third, the response of interest rates to inflation is particularly strong during periods when central bankers want to break a record of high inflation, such as in the UK or Australia at the beginning of the 1980s.
Klíčová slova:
Taylor rule; Inflation targeting; time-varying parameter model
Dokument je dostupný na externích webových stránkách.
How Does Monetary Policy Change? Evidence on Inflation Targeting Countries
We examine the evolution of monetary policy rules in a group of inflation targeting countries (Australia, Canada, New Zealand, Sweden and the United Kingdom), applying a moment-based estimator in a ...
Multifractal height cross-correlation analysis
Krištoufek, Ladislav
2010 - anglický
We introduce a new method for detection of long-range cross-correlations and cross-multifractality – multifractal height cross-correlation analysis (MF-HXA). We show that long-range cross-correlations can be caused by long-range dependence of separate processes and the correlations above them. Similar separation applies for cross-multifractality – standard sep- aration between distributional properties and correlations is enriched by division of correlations between auto-correlations and cross-correlations. Efficiency of the method is showed on two types of simulated series – ARFIMA and Mandelbrot’s Binomial Multifractal model. We further ap- ply the method on returns and volatility of NASDAQ and S&P500 indices and uncover some interesting results.
Klíčová slova:
multifractality; long-range dependence; cross-correlations
Dokument je dostupný na externích webových stránkách.
Multifractal height cross-correlation analysis
We introduce a new method for detection of long-range cross-correlations and cross-multifractality – multifractal height cross-correlation analysis (MF-HXA). We show that long-range cross-correlations ...
Impact of forgetting on models of rolling mills
Dedecius, Kamil; Jirsa, Ladislav
2010 - anglický
The research report deals with an analysis of various models for modelling of the cold sheet rolling process. It comprises a thorough analysis of a mass-flow model and its weaknesses, brief analysis of normalization impact on modelling and exhaustive analysis of 4 defined models with exponential and partial forgetting and their comparison to models without forgetting. The report ends with a computer-intensive search for new blackbox models.
Klíčová slova:
bayesian modelling; forgetting; parameter estimation
Dokument je dostupný na externích webových stránkách.
Impact of forgetting on models of rolling mills
The research report deals with an analysis of various models for modelling of the cold sheet rolling process. It comprises a thorough analysis of a mass-flow model and its weaknesses, brief analysis ...
Divergences between models and data under hypothetical and empirical quantizations
Vajda, Igor; van der Meulen, Edward
2010 - anglický
It is shown that the hypothetical quantization generates classical Pearson-type statistical criteria while the empirical quantization generates criteria asymptotically equivalent to the classical spacings-type statistical criteria. Asymptotic properties of the empirically generated criteria are derived and optimality of the quadratic criterion is proved. Je ukazáno, že hypotetické kvantování generuje clasická statistická kriteria Pearsonova typu, zatim co empirické kvantování generuje kriteria asymptoticky ekvivalentní klasickým přírůstkovým statistickým kriteriím.Jsou odvozeny asymptotické vlastnosti empiricky generovaných kriterií a dokázána optimálnost kvadratického kriteria.
Klíčová slova:
Disparities; Hypothetical quantization; Asymptotic normality; Power divergences; Empirical quantization; Divergences; Asymptotic optimality; Consistency
Dokument je dostupný na externích webových stránkách.
Divergences between models and data under hypothetical and empirical quantizations
It is shown that the hypothetical quantization generates classical Pearson-type statistical criteria while the empirical quantization generates criteria asymptotically equivalent to the classical ...
NUŠL poskytuje centrální přístup k informacím o šedé literatuře vznikající v ČR v oblastech vědy, výzkumu a vzdělávání. Více informací o šedé literatuře a NUŠL najdete na webu služby.
Vaše náměty a připomínky posílejte na email nusl@techlib.cz
Provozovatel
Zahraniční báze