doc. Mgr. Zuzana Hübnerová, Ph.D.
E-mail:   hubnerova@fme.vutbr.cz  Dept.:   Institute of Mathematics Dept. of Statistics and Optimization Position:   Head of Department Room:   A1/1939 Dept.:   Institute of Mathematics Dept. of Statistics and Optimization Position:   Associate Professor Room:   A1/1939

Supervised courses:

• Probability and Statistics II (SP2)
• Probability and Statistics III (SP3)
• Probability and Statistics III (SP3-A)
• Stochastic Processes (SSP)
• Stochastic Processes (SSP-A)

Publications:

• KONEČNÁ, T.; HÜBNEROVÁ, Z.:
  Asymptotic Comparison of Parameters Estimates of Two-parameter Weibull Distribution, Mendel University Press
  journal article in Scopus
• Patrick Langthaler, Yvonne Holler, Zuzana Hubnerova, Vitezslav Vesely and Arne C. Bathke:
  EEG, Nonparametric Multivariate Statistics, and Dementia Classification,
  Statistics and Simulation, IWS 8, Vienna, Austria, September 2015, pp.243-257, ISBN 978-3-319-76034-6, (2018), Springer International Publishing
  book chapter
• TEJKAL, M.; HÜBNEROVÁ, Z.:
  Comparison of Approaches to Testing Equality of Expectations Among Samples from Poisson and Negative Binomial Distribution,
  Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, Vol.66, (2018), No.4, pp.1025-1034, ISSN 1211-8516, Mendel University Press
  journal article - other
• HÜBNEROVÁ, Z.; MICHÁLEK, J.:
  Analysis of daily average PM10 predictions by generalized linear models in Brno, Czech Republic, Turkish National Committee for Air Pollution Research and Control
  journal article in Web of Science
• HÜBNEROVÁ, Z.; DOUDOVÁ, L.:
  Properties of some test of equality of two negative binomial samples,
  AIP Conference Proceedings, pp.1839-1842, ISBN 978-0-7354-1185-2, (2013), AIP Publishing LLC
  conference paper
  akce: ICNAAM 2013 - International Conference on Numerical Analysis and Applied Mathematics, Rhodes, 21.09.2013-27.09.2013
• HÜBNEROVÁ, Z.; VÁŇOVÁ, T.; PROCHÁZKA, J.:
  Estimates of item parameters in computer adaptive testing,
  19th International Conference on Soft Computing MENDEL 2013, pp.421-426, ISBN 978-80-214-4755-4, (2013), FSI VUT Brno
  conference paper
  akce: 19th International Conference on Soft Computing, MENDEL 2013, Brno University of Technology, 26.06.2013-28.06.2013
• HÜBNEROVÁ, Z.; DOUDOVÁ, L.:
  Comparison of Powers of Some Tests in One-Way ANOVA Type Model with Negative Binomial Distribution,
  XX International Conference PDMU-2012, Problems of Decision Making under Uncertainties Proceedings - Applied Papers, pp.19-30, ISBN 978-80-7231-897-1, (2012), Publishing office of the University of Defence
  conference paper
  akce: Problems of Decision Making under Uncertainties PDMU 2012, Brno, 17.09.2012-21.09.2012
• STADLOBER, E.; HÜBNEROVÁ, Z.; MICHÁLEK, J.; KOLÁŘ, M.:
  Forecasting of Daily PM10 Concentrations in Brno and Graz by Different Regression Approaches,
  Austrian Journal of Statistics, Vol.41, (2012), No.4, pp.287-310, ISSN 1026-597X
  journal article - other
• VESELÝ, V.; MICHÁLEK, J.; HÜBNEROVÁ, Z.; KOLÁŘ, M.:
  Vybrané statistické metody pro predikci rizkových jevů,
  Dynamická geovizualizace v krizovém managementu, pp.146-158, ISBN 978-80-210-5858-3, (2011)
  book chapter
• HÜBNEROVÁ, Z.; DOUDOVÁ, L.:
  Influence of Estimate of Shape Parameter on Tests in Models with Negative Binomial Distribution.,
  Biometric Methods and Models in Current Science and Research, pp.119-126, ISBN 978-80-7401-028-6, (2011), ukzuz
  conference paper
  akce: XIX. Letní škola biometriky, Zahradnická fakulta MZLU, Lednice, 06.09.2010-10.09.2010

List of publications at Portal BUT

Abstracts of most important papers:

• HÜBNEROVÁ, Z.; MICHÁLEK, J.:
  Analysis of daily average PM10 predictions by generalized linear models in Brno, Czech Republic, Turkish National Committee for Air Pollution Research and Control
  journal article in Web of Science
  Ambient air quality assessment and management plays an important role in the current European Union policy. Among others, air pollution by PM10 is being monitored. A previous analysis of PM10 (Hrdličková et. al, 2008) aimed at identifying factors affecting air pollution in stations in the City of Brno using available observed meteorological variables. However, the studied model cannot be used for predicting the level of PM10 pollution because the included meteorological variables are not available exactly at the time when the prediction is requested. In that case, we should base the predictions on available predicted variables, namely on temperature, wind direction, wind speed, and cloud cover. A comparison of obtained predictions with the observed values of PM10 during a testing period allows us to evaluate a loss of prediction quality when the predicted covariates are used instead of the observed ones. The presented analysis based on test of symmetry and test of homogeneity of the marginal distribution of two-way contingency tables shows that the loss of prediction quality by employing the predicted meteorological variables is non-significant in the studied case. This observation suggests that the model with predicted meteorological variables can be employed in environmental management processes.
• VESELÝ, V.; TONNER, J.; HRDLIČKOVÁ, Z.; MICHÁLEK, J.; KOLÁŘ, M.:
  Analysis of PM10 Air Pollution In Brno Based on Generalized Linear Model With Strongly Rank-Deficient Design Matrix,
  Environmetrics, Vol.20, (2009), No.6, pp.676-698, ISSN 1180-4009, John Wiley & Sons
  journal article - other
  An analysis of air pollution by suspended particulate matter PM10 in Brno, the second largest urban agglomeration of the Czech Republic, based on generalized linear model (GLM) is presented. Average daily concentrations coming from PM10 monitoring for the period 1998-2005 have been processed. The measured meteorological factors: air temperature and humidity, direction and wind speed were considered as covariates along with some additional seasonal factors. A family of complex (generalized) linear models has been suggested exhibiting strong rank-deficiency in the design matrix to allow for more precise modeling involving identification of significant air pollution sources, among others. From each of them the parameter estimates were obtained using both standard estimation procedure and a new sparse parameter estimation technique based on BPA4 - a four-step modification of the Basis Pursuit Algorithm originally suggested in [Chen S. S., Donoho D. L., Saunders M. A.. Atomic decomposition by basis pursuit. SIAM J. Sci. Comput., 20(1):33--61, 1998.] for time- scale analysis of digital signals. As the standard estimation algorithms often fail due to numerical instability caused by strong overparametrization, we have applied this new computationally intensive approach allowing us to reliably identify nearly zero parameters in the model and thus to find numerically stable sparse solutions. The goal of the analysis was to identify the model and algorithm yielding most precise one-day forecasts of the level of pollution by PM10 with regard to the meteorological and seasonal covariates.
• HRDLIČKOVÁ, Z.:
  Approximation of powers of some tests in one-way MANOVA type multivariate generalized linear model,
  COMPUTATIONAL STATISTICS & DATA ANALYSIS, Vol.52, (2008), No.8, pp.4059-4075, ISSN 0167-9473, Elsevier
  journal article - other
  The test of an equality of expected values of response vectors in one-way MANOVA type multivariate generalized linear models can be based on deviance or score statistic. This paper aims at approximation of the power of these tests under local alternatives. As indicated by presented simulation study, these approximations perform well over a wide range of models and sample sizes. Approximation of the power of these tests based on a noncentral chi-square distribution derived by Fahrmeir [Fahrmeir, L., 1987. Asymptotic testing theory for generalized linear models. Statistics 18 (1), 65-76] is considered for comparison in the simulation study. The approximation derived in this paper performs better for small sample sizes. The simulation study also provides comparison of the considered tests.
• HRDLIČKOVÁ, Z.; MICHÁLEK, J.; KOLÁŘ, M.; VESELÝ, V.:
  Identification of Factors Affecting Air Pollution by Dust Aerosol in Brno City, Czech Republic,
  ATMOSPHERIC ENVIRONMENT, Vol.42, (2008), No.37, pp.8661-8673, ISSN 1352-2310, Elsevier
  journal article - other
  The statistical analysis of the observation of dust aerosol PM10 from four monitoring stations of the agglomeration of the city of Brno during a time period from January 1, 1998 until December 30, 2005 is presented. The main meteorological factors affecting air pollution at each station were identified by means of generalized autoregressive linear model with gamma distribution of response variable and log-link function. Along with meteorological factors, the influence of heating season and weekdays on the air pollution was considered. The suggested model can be used for a prediction of the value of dust aerosol PM10 at given station using selected factors and their previous values.
• HRDLIČKOVÁ, Z.:
  Comparison of the power of the tests in one-way ANOVA type model with Poisson distributed variables,
  Environmetrics, Vol.17, (2006), No.3, pp.227-237, ISSN 1180-4009, John Wiley & Sons, Ltd.
  journal article - other
  akce: 14. International Conference on Quantitative Methods for Environmental Sciences,TIES 2003, Johannesburg, 03.11.2003-07.11.2003
  This article deals with the one-way ANOVA type generalized linear model, the method most often used if the assumptions of the classical one-way ANOVA are not satisfied. In particular, the model with Poisson distributed response variables is considered. The article aims at the powers of tests based on the most frequent statistics - deviance, Pearson and score statistic. The asymptotic powers of the considered tests are obtained either analytically or by means of simulation. Afterwards, these powers are compared with powers of the tests obtained by an alternative method - a variance-stabilizing transformation followed by the classical linear model with known or unknown variance. Finally, a power analysis is carried out on biometrical data to find an experimental design suitable for detecting a priory given differences in expected values.