|
Modelling of land cover change in abandoned landscape using time series of aerial photography
Brůna, Josef ; Jačková, Kateřina (referee) ; Vojta, Jaroslav (advisor)
Obsah 5 English abstract This thesis deals with modelling of land cover change in abandoned landscape. The study site is located in the military area Hradiště, around the former village Tocov, it is 2,5 km long and 2 km wide. Due to lack of historical vegetation data on landscape scale, historical aerial photographs were used as a primary source of data. They were orthorectified, mosaicked and automatically object oriented then classified in two categories woodland and grassland. Changes between two successive classified images were modelled with generalized linear models with mixed effects (lmer). Variables that were derived from digital elevation model, former land use and spatial variables computed with algorithms based on cellular automata were used. In order to verify the model predictions with the actual situation and for easier interpretation and visualization of results, a new application PEMZOK (Spatially explicit model of overgrowth of abandoned landscape) was developed and is one of the results of this work. Rapid rate of overgrowth of remaining grasslands was found. The largest observed effect on the spread of trees had the distance to the nearest tree or shrub and density of trees in the vicinity (15 m × 15 m) and the wider neighbourhood (105 m × 105 m). Models with these variables and land use...
|
|
Portfólio Value at Risk a Expected Shortfall s použitím vysoko frekvenčních dat
Zváč, Marek ; Fičura, Milan (advisor) ; Janda, Karel (referee)
The main objective of this thesis is to investigate whether multivariate models using Highfrequency data provide significantly more accurate forecasts of Value at Risk and Expected Shortfall than multivariate models using only daily data. Our objective is very topical since the Basel Committee announced in 2013 that is going to change the risk measure used for calculation of capital requirement from Value at Risk to Expected Shortfall. The further improvement of accuracy of both risk measures can be also achieved by incorporation of high-frequency data that are rapidly more available due to significant technological progress. Therefore, we employed parsimonious Heterogeneous Autoregression and its asymmetric version that uses high-frequency data for the modeling of realized covariance matrix. The benchmark models are chosen well established DCC-GARCH and EWMA. The computation of Value at Risk (VaR) and Expected Shortfall (ES) is done through parametric, semi-parametric and Monte Carlo simulations. The loss distributions are represented by multivariate Gaussian, Student t, multivariate distributions simulated by Copula functions and multivariate filtered historical simulations. There are used univariate loss distributions: Generalized Pareto Distribution from EVT, empirical and standard parametric distributions. The main finding is that Heterogeneous Autoregression model using high-frequency data delivered superior or at least the same accuracy of forecasts of VaR to benchmark models based on daily data. Finally, the backtesting of ES remains still very challenging and applied Test I. and II. did not provide credible validation of the forecasts.
|
|
Gradient Boosting Machine and Artificial Neural Networks in R and H2O
Sabo, Juraj ; Bašta, Milan (advisor) ; Plašil, Miroslav (referee)
Artificial neural networks are fascinating machine learning algorithms. They used to be considered unreliable and computationally very expensive. Now it is known that modern neural networks can be quite useful, but their computational expensiveness unfortunately remains. Statistical boosting is considered to be one of the most important machine learning ideas. It is based on an ensemble of weak models that together create a powerful learning system. The goal of this thesis is the comparison of these machine learning models on three use cases. The first use case deals with modeling the probability of burglary in the city of Chicago. The second use case is the typical example of customer churn prediction in telecommunication industry and the last use case is related to the problematic of the computer vision. The second goal of this thesis is to introduce an open-source machine learning platform called H2O. It includes, among other things, an interface for R and it is designed to run in standalone mode or on Hadoop. The thesis also includes the introduction into an open-source software library Apache Hadoop that allows for distributed processing of big data. Concretely into its open-source distribution Hortonworks Data Platform.
|
|
Modelování nestlačitelných ionizovaných směsí
Havrda, Jaroslav ; Roubíček, Tomáš (advisor) ; Bulíček, Miroslav (referee) ; Hron, Jaroslav (referee)
In the present work we study the model of incompressible ionized mixtures from point of view of mathematical analysis and practical implementations. The model consists of the continuity equation, Navier-Stokes equations, Nernst-Planck equations and Poisson equation for the electric potential. We proove the existence of a weak solution for the model in three space dimensions. We also present a simple program for computing the solutions of Nernst-Planck equations and Poisson equation and use it to examine the problem that arrise during the computation. The results in thermodynamics of mixtures are refered as the physical part of the thesis.
|
|
Computer simulation of radiobiological effect of oxygen
Havle, Oto ; Knobloch, Petr (referee) ; Felcman, Jiří (advisor)
Exposing a live cell to ionizing radiation can lead to cell inactivation. Oxygen and other chemical substances soluted in the internal environment of a cell participate in various chemical reactions during the chemical stage of the radiobiological process. The thesis is concerned with modelling di usion and reaction processes of the chemical stage. A system of semilinear parabolic partial di erential equations for concentrations of chemical substances involved is derived in three spatial dimensions, including weak formulation. Assuming spherical symmetry, the problem can be reduced to one dimension. Under further simpli cations, existence of solution is proven via Galerkin method. The one-dimensional problem is solved numerically, using nite element discretization. Error estimates and computer implementation are presented.
|
|
The dehumanization of thinking
Kršňák, Jan ; Hanáková, Petra (referee) ; Petříček, Miroslav (advisor)
This is not a summary, but a meaningless part of my thesis. Even it is not a part of it at all. There is no summary in czech. There is no method, but a lot of ruptures, errors, guessworks, and above all gaps. This text is a singular discourse, closed into itself like a black hole and a conceptual film, which has to be seen by reading and read by looking. There is no particular message, this storytelling has no start and no end and doesn't care about present anything to a reader. It is a text which has to be used, because you can't find out any meanings here, if you don't participate in creating them. It is a pulse burst, acummulation of small stories. I use biology, physics, geology, bacteriology and other siences that I don't understand, what helps me to use their stories by the way I need to tell my story which consists only of gaps. There is no diference between physical laws and rules of football. Nature has no grammar and some particles deny to be right, others are missing. Human brain is a result of an easy level of computer game. I don't read Aristotle because of he hadn't a TV set at home. Racionality was dissolved by storytelling and imagination. I use works of Deleuze, Guattari, Baudrillard, Virilio, Debord, De Landa and advertising industry. A rock is considered to be a movie here. A landscape is...
|
| |
|
Modeling of plants using "Point-sprites"
Klecanda, Václav ; Pelikán, Josef (advisor) ; Mráz, František (referee)
In the present work i study possibilities of using elemets of hardware accelerated computer graphics called Point sprites in modeling and rendering of natural objects as trees, bushes, flowers. Next goal of the work is possibility of connecting using Point sprites with one of commonly used technique LoD( Level of Detail) for achieving best effectivity of rendering.
|
|
Energy evaluation of office building.
Lysková, Markéta ; Horák, Petr (referee) ; Adam, Pavel (advisor)
This diploma thesis is dealing with energy audit of industrial building with administration part, situated in Blansko, South Moravia region. The theoretical part is focused to windows from the energy balance point of view, because one of the energy saving measures is the change of current insufficient window filling. The main objective of this diploma thesis is to find the most energy saving variant which is going to be evaluated from the both, ecological and economical, aspect. The third part is dedicated to usage of computer technology as software for 3D model creating of assessed building or energy consumption simulation for particular systems as heating in relation to characters of different kinds of windows.
|
|
Steel structure of a business centre
Kočí, Petr ; Pilgr, Milan (referee) ; Štrba, Michal (advisor)
The diploma thesis elaborates on the proposal of a loadbearing steel structure of a shopping centre. Four-storey building has total lenght of 88 m, width 48 m and its base structural grid is 8 x 8 m. The total height of building is 25,4 m. First three floors above ground are designed with structural height of 5,6 m whilst the 4TH floor uses height of 4,0 m. Frame structure is formed by primary beams and castellated secondary beams. Above 4TH floor, the roof structure is designed in two alternative versions, both using solid purlins and truss girder. The 3D computer animation of the building is also part of thesis. The shopping centre that is subject of the final thesis is located in Brno area.
|
guest ::
