|
Clustering and regression analysis of micro panel data
Sobíšek, Lukáš ; Pecáková, Iva (advisor) ; Komárek, Arnošt (referee) ; Brabec, Marek (referee)
The main purpose of panel studies is to analyze changes in values of studied variables over time. In micro panel research, a large number of elements are periodically observed within the relatively short time period of just a few years. Moreover, the number of repeated measurements is small. This dissertation deals with contemporary approaches to the regression and the clustering analysis of micro panel data. One of the approaches to the micro panel analysis is to use multivariate statistical models originally designed for crosssectional data and modify them in order to take into account the within-subject correlation. The thesis summarizes available tools for the regression analysis of micro panel data. The known and currently used linear mixed effects models for a normally distributed dependent variable are recapitulated. Besides that, new approaches for analysis of a response variable with other than normal distribution are presented. These approaches include the generalized marginal linear model, the generalized linear mixed effects model and the Bayesian modelling approach. In addition to describing the aforementioned models, the paper also includes a brief overview of their implementation in the R software. The difficulty with the regression models adjusted for micro panel data is the ambiguity of their parameters estimation. This thesis proposes a way to improve the estimations through the cluster analysis. For this reason, the thesis also contains a description of methods of the cluster analysis of micro panel data. Because supply of the methods is limited, the main goal of this paper is to devise its own two-step approach for clustering micro panel data. In the first step, the panel data are transformed into a static form using a set of proposed characteristics of dynamics. These characteristics represent different features of time course of the observed variables. In the second step, the elements are clustered by conventional spatial clustering techniques (agglomerative clustering and the C-means partitioning). The clustering is based on a dissimilarity matrix of the values of clustering variables calculated in the first step. Another goal of this paper is to find out whether the suggested procedure leads to an improvement in quality of the regression models for this type of data. By means of a simulation study, the procedure drafted herein is compared to the procedure applied in the kml package of the R software, as well as to the clustering characteristics proposed by Urso (2004). The simulation study demonstrated better results of the proposed combination of clustering variables as compared to the other combinations currently used. A corresponding script written in the R-language represents another benefit of this paper. It is available on the attached CD and it can be used for analyses of readers own micro panel data.
|
|
Motivation and evaluation of empolyees
Sinkulová, Anna ; Legnerová, Kateřina (advisor) ; Šmíd, David (referee)
My bachelor thesis is focused on motivation and assessment of employees in the company. The aim of this thesis is to investigate the influence of team building and benefits on employees' motivation. Literature based theoretical part of my thesis describes and investigates the real meaning and purpose of motivation and basic motivational concepts. Furthermore, I outlined the basics behind employee benefits, team building and also stated the current given trends in these areas. The practical part is focused on benefit and workplace research and their impact on the general motivation to work and the relations between the employees of the Czechoslovak Commercial Bank (ČSOB). In order to collect the best results, I used the method of questionnaires. Respondents were asked to answer questions considering their satisfaction with the system of employee benefits and the importance of teamwork. Based on the survey results I propose certain changes that should lead to motivation improvement and better job performance.
|
|
Progress, outcome, future scenarios and threats of multilateral trading system (WTO)
Kozáková, Michaela ; Štěrbová, Ludmila (advisor) ; Trojanová, Kamila (referee)
This diploma thesis aims to set a complex view on the current process of multilateral trade agreements under the World Trade Organization and point out the direction of DOHA negotiations. The emphasis is put on the Tenth Ministerial Conference, which took place in Nairobi in December 2015. Stances and demands of the WTO member are analysed step by step. The thesis considers threats that possibly come from plurilateral, bilateral and regional agreements and evaluate a potential risk for the multilateral trading system. For this purpose, the thesis is divided into four chapters. First chapter examine theoretical introduction to WTO and particularly to Development Doha Agenda, which is important for subsequent understanding of actual issues. Second part analyses in detail the negotiations in main negotiating parts of DDA before and after a summer break. Following chapter fluently continues with analyse of the progress after summer break and points out some current questions about preparations for MC10 and expectations. Space is also given to the MC10 and its outcomes. Finally, the last chapter gives a thought to future scenarios concerning function of the WTO.
|
|
ELECTRON BEAM REMELTING OF PLASMA SPRAYED ALUMINA COATINGS
Matějíček, Jiří ; Veverka, J. ; Čížek, J. ; Kouřil, J.
Plasma sprayed alumina coatings find numerous applications in various fields, where they enhance the properties of the base material. Examples include thermal barriers, wear resistance, electrical insulation, and diffusion and corrosion barriers. A typical structure of plasma sprayed coatings, containing a multitude of voids and imperfectly bonded interfaces, gives them unique properties - particularly low thermal conductivity, high strain tolerance, etc. However, for certain applications such as permeation barriers or wear resistance, these voids may be detrimental.\nThis paper reports on the first experiments with remelting of plasma sprayed alumina coatings by electron beam technology, with the purpose of densifying the coatings and thereby eliminating the voids. Throughout the study, several parameters of the e-beam device were varied - beam current, traverse velocity and number of passes. The treated coatings were observed by light and electron microscopy and the thickness, structure and surface morphology of the remelted layer were determined and correlated with the process parameters. Based on the first series of experiments, the e-beam settings leading to dense and smooth remelted layer of sufficient thickness were obtained. In this layer, a change of phase composition and a marked increase in hardness were observed.\n
|
|
Travel Behaviour of the Prague Suburb Population
Pergl, Ondřej ; Novák, Jakub (advisor) ; Marada, Miroslav (referee)
Travel behavior of Prague suburb population Abstract The hinterland of Prague is formed by the process of suburbanization, which influences the spatial structure of whole region. Changed spatial structure puts greater demand on the mobility of the population. The main aim of this work is the research of travel behavior of the Prague suburb population, who makes his movements primarily by the car. The work focuses on the factors influencing travel behavior, which are discussed in theoretical section. The next section discusses the measures against individual transport. The empirical part presents the results of survey in Prague southeast hinterland. Travel behavior is examined in terms of modal split and trip purpose. Further investagation is focused on citizens' satisfaction with the traffic situation in their municipality and reveals their sensitivity to various measures against car use. Key words: mobility, travel behavior, suburbanization, car use limiting, Prague
|
|
Effect of snowpack on runoff generation during rain on snow event.
Juras, Roman ; Máca, Petr (advisor) ; Ladislav , Ladislav (referee)
During a winter season, when snow covers the watershed, the frequency of rain-on-snow (ROS) events is still raising. ROS can cause severe natural hazards like floods or wet avalanches. Prediction of ROS effects is linked to better understanding of snowpack runoff dynamics and its composition. Deploying rainfall simulation together with hydrological tracers was tested as a convenient tool for this purpose. Overall 18 sprinkling experiments were conducted on snow featuring different initial conditions in mountainous regions over middle and western Europe.
Dye tracer brilliant blue (FCF) was used for flow regime determination, because it enables to visualise preferential paths and layers interface. Snowpack runoff composition was assessed by hydrograph separation method, which provided appropriate results with acceptable uncertainty. It was not possible to use concurrently these two techniques because of technical reasons, however it would extend our gained knowledge. Snowmelt water amount in the snowpack runoff was estimated by energy balance (EB) equation, which is very efficient but quality inputs demanding. This was also the reason, why EB was deployed within only single experiment.
Timing of snowpack runoff onset decrease mainly with the rain intensity. Initial snowpack properties like bulk density or wetness are less important for time of runoff generation compared to the rain intensity. On the other het when same rain intensity was applied, non-ripe snowpack featuring less bulk density created runoff faster than the ripe snowpack featuring higher bulk density. Snowpack runoff magnitude mainly depends on the snowpack initial saturation. Ripe snowpack with higher saturation enabled to generate higher cumulative runoff where contributed by max 50 %. In contrary, rainwater travelled through the non-ripe snowpack relatively fast and contributed runoff by approx. 80 %.
Runoff prediction was tested by deploying Richards equation included in SNOWPACK model. The model was modified using a dual-domain approach to better simulate snowpack runoff under preferential flow conditions. Presented approach demonstrated an improvement in all simulated aspects compared to the more traditional method when only matrix flow is considered.
|
|
Individual human odor as a forensic trail in criminal proceeding made by contact or contactless transmission and its resistance to physical agents
Santariová, Milena ; Bartoš, Luděk (advisor) ; Chmelíková, Eva (referee)
The dissertation thesis consists of four papers that have been published in scientific journals.
Study n. 1
The need to recover evidence from water is quite common in criminal investigation. The article deals with the possibility to collect human scent from such objects and with the ability of specially trained dogs to match such scent samples with scent samples collected from detained suspects. During an experiment, designed as a blind one, it has been proved, that human scent can survive on submerged objects and can be later used for scent identification.
Study n. 2
To collect odors the Czech Republic Police use special fabric sorbents manufactured under the registered mark ARATEX. Before use the fabric sorbent is treated by water vapor sterilization. After the scent identification the sorbent is destroyed. The goal of the study was to verify if the vapor sterilization is effective enough to remove human scent that has already penetrated into the sorbent structure or in other words if the sorbent can be exposed to vapor sterilization and then used again. Specially trained dogs were used to match starting odors with target odors in the line-ups. The results showed that dogs are able to correctly match human odors even after they have being exposed to vapor sterilization.
Study n. 3
The purpose of the study was to determine the temperature at which the human scent is degraded so that a dog would not be able to identify it. In contrast to expectations, eight dogs used in the experiment almost flawlessly identified human scents from five scent donors exposed to temperatures of 100°C, 200°C, 300°C, 400°C, 500°C, 600°C, 700°C, and 800°C. Only two of the dogs were able to identify 5 of 15 scent samples exposed to 900°C. No dog identified a scent exposed to 1000°C. Our study verified heat survivability of human scent far beyond existing expectations.
Study n. 4
We tested the hypothesis that if odor fallout (the release of a human odor onto an untouched object) in human subjects exists, then holding a hand above an absorbent will produce a detectable scent which will be subsequently matched in a detection test by trained dogs. Scents were collected from seven males to sterile cotton absorbent squares. The left hand was used to get the control scent and the right hand served as the target scent. Each experimental subject was sitting and his left hand was laid down on a cotton square for 3 min. The right hand was held 5 cm above another cotton square for 3 min. The scent identification was done by two specially trained police German shepherds. Both dogs performed 14 line-ups and correctly matched the collected scents of all test subjects. The results suggest the existence of human odor fallout, whereby a human scent trace is left by humans even if they do not touch an object.
|
|
Non-invasive methods for sex and steroid homones determination in gibbons of the genus Nomascus
Bolechová, Petra ; Bartoš, Luděk (advisor) ; Pavel, Pavel (referee)
The gibbon primates of the family Nomascus are classified as critically endangered species, and, to date, basic understanding and information about their biology is missing. With regard to the status of these animals in the wild and captive populations in zoos, being familiar with their reproduction, may improve captive breeding programs. Data collection in this study was to be carried out by the practical use of non-invasive methods (polymerase chain reaction for DNA extraction and enzyme immunoassays), using faecal samples for analysis. The first method, polymerase chain reaction, was used for sex determination in juveniles, because of their coat colour and visual similarity of secondary sex characteristic in both sexes; it is not possible to determinate sex without handling the animal. Another main purpose of this study is to try and answer the hypothesis regarding the ovarian cycle of females, factors influencing their hormone concentration and also the onset of sexual maturity in females and the timing of their fur colour change. Hypotheses were checked by monitoring the concentration of progesterone and oestrogen faecal metabolites and by evaluation of the composition of breeding groups of gibbons with the ZOO influence. During a four year period (from 2010 till 2014), there were a total of 51 animals analysed from 16 different zoos with a faecal sample count totalling 1618 samples. The results confirmed the use of noninvasive methods for sex determination, and thus ensuring the maximum welfare standards. Endocrinological analysis confirmed the hypothesis of the influence of the environment (ZOO) to the hormone concentrations and female´s pregnancy effect in conjunction with her age on the final results of the male - father hormone concentrations. The initial information in this study is the confirmation of the onset of ovarian cycle in young females without connection to their fur colour change and a significant factor of a mother´s impact and her territoriality. This study is the first to present a long term monitoring of ovarian cycles in females and hormone concentrations of other individuals, both male and female, and from various age groups. The results allow us to understand the possible impact of zoo environments on the reproductive status of these gibbons and contribute to the general improvement of breeding management.
|
|
New Methods for Increasing Efficiency and Speed of Functional Verification
Zachariášová, Marcela ; Dohnal, Jan (referee) ; Steininger, Andreas (referee) ; Kotásek, Zdeněk (advisor)
Při vývoji současných číslicových systémů, např. vestavěných systému a počítačového hardware, je nutné hledat postupy, jak zvýšit jejich spolehlivost. Jednou z možností je zvyšování efektivity a rychlosti verifikačních procesů, které se provádějí v raných fázích návrhu. V této dizertační práci se pozornost věnuje verifikačnímu přístupu s názvem funkční verifikace. Je identifikováno několik výzev a problému týkajících se efektivity a rychlosti funkční verifikace a ty jsou následně řešeny v cílech dizertační práce. První cíl se zaměřuje na redukci simulačního času v průběhu verifikace komplexních systémů. Důvodem je, že simulace inherentně paralelního hardwarového systému trvá velmi dlouho v porovnání s během v skutečném hardware. Je proto navrhnuta optimalizační technika, která umisťuje verifikovaný systém do FPGA akcelerátoru, zatím co část verifikačního prostředí stále běží v simulaci. Tímto přemístěním je možné výrazně zredukovat simulační režii. Druhý cíl se zabývá ručně připravovanými verifikačními prostředími, která představují výrazné omezení ve verifikační produktivitě. Tato režie však není nutná, protože většina verifikačních prostředí má velice podobnou strukturu, jelikož využívají komponenty standardních verifikačních metodik. Tyto komponenty se jen upravují s ohledem na verifikovaný systém. Proto druhá optimalizační technika analyzuje popis systému na vyšší úrovni abstrakce a automatizuje tvorbu verifikačních prostředí tím, že je automaticky generuje z tohoto vysoko-úrovňového popisu. Třetí cíl zkoumá, jak je možné docílit úplnost verifikace pomocí inteligentní automatizace. Úplnost verifikace se typicky měří pomocí různých metrik pokrytí a verifikace je ukončena, když je dosažena právě vysoká úroveň pokrytí. Proto je navržena třetí optimalizační technika, která řídí generování vstupů pro verifikovaný systém tak, aby tyto vstupy aktivovali současně co nejvíc bodů pokrytí a aby byla rychlost konvergence k maximálnímu pokrytí co nejvyšší. Jako hlavní optimalizační prostředek se používá genetický algoritmus, který je přizpůsoben pro funkční verifikaci a jeho parametry jsou vyladěny pro tuto doménu. Běží na pozadí verifikačního procesu, analyzuje dosažené pokrytí a na základě toho dynamicky upravuje omezující podmínky pro generátor vstupů. Tyto podmínky jsou reprezentovány pravděpodobnostmi, které určují výběr vhodných hodnot ze vstupní domény. Čtvrtý cíl diskutuje, zda je možné znovu použít vstupy z funkční verifikace pro účely regresního testování a optimalizovat je tak, aby byla rychlost testování co nejvyšší. Ve funkční verifikaci je totiž běžné, že vstupy jsou značně redundantní, jelikož jsou produkovány generátorem. Pro regresní testy ale tato redundance není potřebná a proto může být eliminována. Zároveň je ale nutné dbát na to, aby úroveň pokrytí dosáhnutá optimalizovanou sadou byla stejná, jako u té původní. Čtvrtá optimalizační technika toto reflektuje a opět používá genetický algoritmus jako optimalizační prostředek. Tentokrát ale není integrován do procesu verifikace, ale je použit až po její ukončení. Velmi rychle odstraňuje redundanci z původní sady vstupů a výsledná doba simulace je tak značně optimalizována.
|
|
Packet Classification Algorithms
Puš, Viktor ; Lhotka,, Ladislav (referee) ; Dvořák, Václav (advisor)
Tato práce se zabývá klasifikací paketů v počítačových sítích. Klasifikace paketů je klíčovou úlohou mnoha síťových zařízení, především paketových filtrů - firewallů. Práce se tedy týká oblasti počítačové bezpečnosti. Práce je zaměřena na vysokorychlostní sítě s přenosovou rychlostí 100 Gb/s a více. V těchto případech nelze použít pro klasifikaci obecné procesory, které svým výkonem zdaleka nevyhovují požadavkům na rychlost. Proto se využívají specializované technické prostředky, především obvody ASIC a FPGA. Neméně důležitý je také samotný algoritmus klasifikace. Existuje mnoho algoritmů klasifikace paketů předpokládajících hardwarovou implementaci, přesto však tyto přístupy nejsou připraveny pro velmi rychlé sítě. Dizertační práce se proto zabývá návrhem nových algoritmů klasifikace paketů se zaměřením na vysokorychlostní implementaci ve specializovaném hardware. Je navržen algoritmus, který dělí problém klasifikace na jednodušší podproblémy. Prvním krokem je operace vyhledání nejdelšího shodného prefixu, používaná také při směrování paketů v IP sítích. Tato práce předpokládá využití některého existujícího přístupu, neboť již byly prezentovány algoritmy s dostatečnou rychlostí. Následujícím krokem je mapování nalezených prefixů na číslo pravidla. V této části práce přináší vylepšení využitím na míru vytvořené hashovací funkce. Díky použití hashovací funkce lze mapování provést v konstantním čase a využít při tom pouze jednu paměť s úzkým datovým rozhraním. Rychlost tohoto algoritmu lze určit analyticky a nezávisí na počtu pravidel ani na charakteru síťového provozu. S využitím dostupných součástek lze dosáhnout propustnosti 266 milionů paketů za sekundu. Následující tři algoritmy uvedené v této práci snižují paměťové nároky prvního algoritmu, aniž by ovlivňovaly rychlost. Druhý algoritmus snižuje velikost paměti o 11 % až 96 % v závislosti na sadě pravidel. Nevýhodu nízké stability odstraňuje třetí algoritmus, který v porovnání s prvním zmenšuje paměťové nároky o 31 % až 84 %. Čtvrtý algoritmus kombinuje třetí algoritmus se starším přístupem a díky využití několika technik zmenšuje paměťové nároky o 73 % až 99 %.
|
guest ::
