Highly Robust Estimation of the Autocorrelation Coefficient
Kalina, Jan; Vlčková, Katarína
2014 - English The classical autocorrelation coefficient estimator in the time series context is very sensitive to the presence of outlying measurements in the data. This paper proposes several new robust estimators of the autocorrelation coefficient. First, we consider an autoregressive process of the first order AR(1) to be observed. Robust estimators of the autocorrelation coefficient are proposed in a straightforward way based on robust regression. Further, we consider the task of robust estimation of the autocorrelation coefficient of residuals of linear regression. The task is connected to verifying the assumption of independence of residuals and robust estimators of the autocorrelation coefficient are defined based on the Durbin-Watson test statistic for robust regression. The main result is obtained for the implicitly weighted autocorrelation coefficient with small weights assigned to outlying measurements. This estimator is based on the least weighted squares regression and we exploit its asymptotic properties to derive an asymptotic test that the autocorrelation coefficient is equal to 0. Finally, we illustrate different estimators on real economic data, which reveal the advantage of the approach based on the least weighted squares regression. The estimator turns out to be resistant against the presence of outlying measurements. Keywords: time series; autoregressive process; linear regression; robust econometrics Available on request at various institutes of the ASCR Highly Robust Estimation of the Autocorrelation Coefficient

The classical autocorrelation coefficient estimator in the time series context is very sensitive to the presence of outlying measurements in the data. This paper proposes several new robust estimators ...

Kalina, Jan; Vlčková, Katarína
Ústav informatiky, 2014

Representations of Boolean Functions by Perceptron Networks
Kůrková, Věra
2014 - English Limitations of capabilities of shallow perceptron networks are investigated. Lower bounds are derived for growth of numbers of units and sizes of output weights in networks representing Boolean functions of d variables. It is shown that for large d, almost any randomly chosen Boolean function cannot be tractably represented by shallow perceptron networks, i.e., each its representation requires a network with number of units or sizes of output weights depending on d exponentially Keywords: perceptron networks; model complexity; Boolean functions Available in digital repository of the ASCR Representations of Boolean Functions by Perceptron Networks

Limitations of capabilities of shallow perceptron networks are investigated. Lower bounds are derived for growth of numbers of units and sizes of output weights in networks representing Boolean ...

Kůrková, Věra
Ústav informatiky, 2014

Prediction diagnostics for Uncertain Systems
Novák, M.; Votruba, Z.; Brandejský, T.; Faber, J.; Coufal, David; Pelikán, Emil
2014 - English Available at various institutes of the ASCR Prediction diagnostics for Uncertain Systems

Novák, M.; Votruba, Z.; Brandejský, T.; Faber, J.; Coufal, David; Pelikán, Emil
Ústav informatiky, 2014

A Weather Risk Prediction System for Road Trip Planning
Krč, Pavel; Fuglík, Viktor; Juruš, Pavel; Kasanický, Ivan; Konár, Ondřej; Pelikán, Emil; Eben, Kryštof; Šucha, M.
2014 - English The paper presents first ideas of the MEDARD-RODOS project. The aim of the project is to develop a decision support system for road trip planning, reflecting the weather risks predicted from the NWP models implemented in the MEDARD system (www.medard-online.cz) and using the traffic information from the RODOS project (www.centrum-rodos.cz). Keywords: weather; prediction; risk; system; road; planning; MEDARD; RODOS; NWP Available on request at various institutes of the ASCR A Weather Risk Prediction System for Road Trip Planning

The paper presents first ideas of the MEDARD-RODOS project. The aim of the project is to develop a decision support system for road trip planning, reflecting the weather risks predicted from the NWP ...

Krč, Pavel; Fuglík, Viktor; Juruš, Pavel; Kasanický, Ivan; Konár, Ondřej; Pelikán, Emil; Eben, Kryštof; Šucha, M.
Ústav informatiky, 2014

Noise revealing in Golub-Kahan bidiagonalization as a mean of regularization in discrete inverse problems
Kubínová, Marie; Hnětynková, Iveta
2014 - English Keywords: ill-posed problems; regularization; Krylov subspace Available in a digital repository NRGL Noise revealing in Golub-Kahan bidiagonalization as a mean of regularization in discrete inverse problems

Kubínová, Marie; Hnětynková, Iveta
Ústav informatiky, 2014

On three equivalent methods for parameter estimation problem based on spatio-temporal FRAP data
Matonoha, Ctirad; Papáček, Š.
2014 - English Keywords: inverse problem formulation; Tikhonov regularizaton; least-squares problem Available in a digital repository NRGL On three equivalent methods for parameter estimation problem based on spatio-temporal FRAP data

Matonoha, Ctirad; Papáček, Š.
Ústav informatiky, 2014

The use of Score Functions of Distribution for Description of Parametric Families
Fabián, Zdeněk
2014 - English Keywords: systems of distributions; Johnson transformations; score function of distribution; parametric families Available on request at various institutes of the ASCR The use of Score Functions of Distribution for Description of Parametric Families

Fabián, Zdeněk
Ústav informatiky, 2014

Explaining Anomalies with Sapling Random Forests
Pevný, T.; Kopp, Martin
2014 - English The main objective of anomaly detection algorithms is finding samples deviating from the majority. Although a vast number of algorithms designed for this already exist, almost none of them explain, why a particular sample was labelled as an anomaly. To address this issue, we propose an algorithm called Explainer, which returns the explanation of sample’s differentness in disjunctive normal form (DNF), which is easy to understand by humans. Since Explainer treats anomaly detection algorithms as black-boxes, it can be applied in many domains to simplify investigation of anomalies. The core of Explainer is a set of specifically trained trees, which we call sapling random forests. Since their training is fast and memory efficient, the whole algorithm is lightweight and applicable to large databases, datastreams, and real-time problems. The correctness of Explainer is demonstrated on a wide range of synthetic and real world datasets. Keywords: anomaly explanation; decision trees; feature selection; random forest Available in digital repository of the ASCR Explaining Anomalies with Sapling Random Forests

The main objective of anomaly detection algorithms is finding samples deviating from the majority. Although a vast number of algorithms designed for this already exist, almost none of them explain, ...

Pevný, T.; Kopp, Martin
Ústav informatiky, 2014

Case Study in Approaches to the Classification of Audiovisual Recordings of Lectures and Conferences
Pulc, P.; Holeňa, Martin
2014 - English Several methods for classification of semistructured documents already exist, thus also classifications for individual modalities of multimedia content. However, every classifier can behave differently on different data modalities and can be differently appropriate for classification of the considered multimedia content as a whole. Because of that, relying on a single classifier or a static weighting of the classification of individual modalities is not adequate. The present paper describes a case study in searching for suitable classification methods, and in investigating appropriate methods for the aggregation of their results to determine a final class of a lecture or conference recording. Keywords: multimedial data; classification; ensembles of classifiers Available in digital repository of the ASCR Case Study in Approaches to the Classification of Audiovisual Recordings of Lectures and Conferences

Several methods for classification of semistructured documents already exist, thus also classifications for individual modalities of multimedia content. However, every classifier can behave ...

Pulc, P.; Holeňa, Martin
Ústav informatiky, 2014

Meta-Parameters of Kernel Methods and Their Optimization
Vidnerová, Petra; Neruda, Roman
2014 - English In this work we deal with the problem of metalearning for kernel based methods. Among the kernel methods we focus on the support vector machine (SVM), that have become a method of choice in a wide range of practical applications, and on the regularization network (RN) with a sound background in approximation theory. We discuss the role of kernel function in learning, and we explain several search methods for kernel function optimization, including grid search, genetic search and simulated annealing. The proposed methodology is demonstrated on experiments using benchmark data sets. Keywords: kernel methods; metalearning; computational intelligence Available in digital repository of the ASCR Meta-Parameters of Kernel Methods and Their Optimization

In this work we deal with the problem of metalearning for kernel based methods. Among the kernel methods we focus on the support vector machine (SVM), that have become a method of choice in a wide ...

Vidnerová, Petra; Neruda, Roman
Ústav informatiky, 2014