Lidé na fakultě
Ing. Šárka Jozová, Ph.D.

2022, Neural Network World, 32 (1), p. 15-41), ISSN 1210-0552

The paper deals with the task of modeling discrete questionnaire data with a reduced dimension of the model. The discrete model dimension is reduced using the construction of local models based on independent binomial mixtures estimated with the help of recursive Bayesian algorithms in the combination with the naive Bayes technique. The main contribution of the paper is a three-phase algorithm of the discrete model dimension reduction, which allows to model high-dimensional questionnaire data with high number of explanatory variables and their possible realizations. The proposed general solution is applied to the traffic accident questionnaire analysis, where it takes the form of the classification of the accident circumstances and prediction of the traffic accident severity using the currently measured discrete data. Results of testing the obtained model on real data and comparison with theoretical counterparts are demonstrated.

Článek v periodiku excerpovaném SCI Expanded

2021, Proceedings of the 18th International Conference on Informatics in Control, Automation and Robotics - ICINCO, Setùbal, SciTePress), p. 641-648), ISBN 978-989-758-522-7, ISSN 2184-2809

This paper aims at presenting the on-line non-iterative form of Bayesian mixture estimation. The model used is composed of a set of sub-models (components) and an estimated pointer variable that currently indicates the active component. The estimation is built on an approximated Bayes rule using weighted measured data. The weights are derived from the so called proximity of measured data entries to individual components. The basis for the generation of the weights are integrated likelihood functions with the inserted point estimates of the component parameters. One of the main advantages of the presented data analysis method is a possibility of a simple incorporation of the available prior knowledge. Simple examples with a programming code as well as results of experiments with real data are demonstrated. The main goal of this paper is to provide clear description of the Bayesian estimation method based on the approximated likelihood functions, called proximities.

Stať ve sborníku z prestižní konf. (Scopus)

2021, Neural Network World, 31 (5), p. 311-328), ISSN 1210-0552

An analysis of survey data is a fundamental part of research concerning various aspects of human behavior. Such survey data are often discrete, and the size of the collected sample is regularly insufficient for the most potent modelling tools such as logistic regression, clustering, and other data mining techniques. In this paper, we take a closer look at the results of the stated preference survey analyzing how inhabitants of cities in Thailand, Poland, and Czechia understand and perceive "smartness" of a city. An international survey was conducted, where respondents were asked 15 questions. Since the most common data modelling tools failed to provide a useful insight into the relationship between variables, so-called lambda coefficient was used and its usefulness for such challenging data was verified. It uses the principle of conditional probability and proves to be truly useful even in data sets with relatively small sample size.

Článek v periodiku excerpovaném SCI Expanded

2021, Neural Network World, 31 (3), p. 227-238), ISSN 1210-0552

According to the statistics about vehicle accidents, there are many causes such as traffic violations, reduced concentration, micro sleep, hasty aggression, but the most frequent cause of accidents at highways is a carelessness of the driver and violation of keeping a safe distance. Producers of vehicles try to take into account this situation by development of assistance systems which are able to avoid accidents or at least to mitigate its consequences. This urgent situation leaded to the described project of investigation of behavior of drivers in dangerous situations occurring in vehicle driving. The research is to help in solution of the present unsatisfactory situation in driving accidents. The developed decisionmaking algorithm of detection serious driving situations that can lead to accidents was tested in the laboratory of driving simulators in FTS CTU, Prague. The data for its testing resembled highway traffic.

Článek v periodiku excerpovaném SCI Expanded

2021, 2021 Smart City Symposium Prague, Piscataway, IEEE Signal Processing Society), ISBN 9780738131580

Data analysis is very important tool for acquiring information about object of our interest. It involves a lot of various methods, mainly from the area of data mining or statistical analysis. However, most of these methods aim at continuous data. In real applications, especially in the field of Smart Cities, questionnaires are a frequent source of data. They are mainly a source of discrete data. As commonly used data analysis methods, such as regression analysis, naturally work with continuous data, complications can occur. However, the linear regression analysis can be carefully used to to analyze discrete data, but carefully.The paper wants to give a warning before a direct mindless use of regression analysis for discrete data, especially when the independent variables are nominal. Also some ways how to modify values of the independent variables to achieve sensible results with linear regression applied to measured discrete data from the Smart City area are sketched.

Stať ve sborníku z mezinár. konf.

2020, 2020 Smart City Symposium Prague, New York, IEEE Press), ISBN 978-1-7281-6821-0

Data analysis is an important method for obtaining information which is useful in many projects such as e.g. Smart Cities. Common data sources are questionnaires, their output mainly purveys discrete data. The most common description of discrete data is through categorical models. These models have several advantages such as flexibility but there are also disadvantages such as a huge dimension of the table expressing this distribution for more variables and values and their overparametrization. The aim of this paper is to replace the categorical distribution by another discrete distribution with a lower number of parameters while maintaining the model quality. Binomial distribution was chosen a suitable one because it is determined only by one parameter and this parameter allows to shape the probability function of binomial distribution well. The output of the paper is the presented model of mixture with binomial components. The suggested estimation algorithms are tested on real traffic data.

Stať ve sborníku z mezinár. konf.

2020, Automa, 26 (1), p. 26-31), ISSN 1210-9592

Článek se zabývá modelováním diskrétních dopravních dat pomocí binomického, geometrického, Poissonova a rovnoměrného rozdělení.

Článek v odborném recenzovaném periodiku česky

2020

In the scope of the Motostudent competition there was created complex project of the whole motorcycle in the category of Electro bikes. Report economical assessemen, industrial production plan

Technická zpráva cizojazyčně

2018

In the scope of the Motostudent competition there was created complex project of the whole motorcycle in the category of Electro bikes. Report contains conceptual design, detailed design, prototyping, economical assessemen, industrialproduction plan and technological innovation.

Oponovaná technická zpráva cizojazyčně

2018

In the scope of the Motostudent competition there was created complex project of the whole motorcycle in the category of Petrol bikes. Report contains conceptual design, detailed design, prototyping, economical assessemen, industrialproduction plan and technological innovation.

Oponovaná technická zpráva cizojazyčně