The diploma theses focused on the topic of removing spectral aberration of sensor. The anomaly is being called spectral smile effect, or spectral curvature. Usually are with this artifact affected hyperspectral sensors, whose sensing is based on principle corresponding to pushbroom scanner. The defect occurs in the data as brightness gradient in the cross-track direction to the sensor flight and affects the object's spectral characteristics shape. Removing is important particularly in case of applying atmospheric correction algorithm, which are being degraded with presence of this defect. The work contents description of current methods used for detecting and eliminating this anomaly. The work is closer focused on the methods removing the defect based on own measured data without use of calibration targets. There is being examined sensitivity of the methods on the scene perception character, meaning heterogeneity of the scene and other outside influence. One of the methods is deeply processed resulting in increasing quality of the method for removing the defect.

Common mistakes Czech students make in English represent the main topic of the presented master thesis. The theoretical part mainly deals with the presentation of various TEFL (Teaching English as Foreign Language) methods. Moreover, they are discussed from several perspectives, including techniques, strategies and principles while using the particular method, the role of the teacher, his or her relationship with students, and especially the response to errors within each of the methods. The purpose of presenting TEFL methods is to show also the connection between mistakes and methods themselves. Concerning the area of mistakes and their corrections, we focus on the classification of mistakes and their division. The practical part is based on various illustrations of mistakes from students' writing assignments. These mistakes are divided into four major classes, including grammar, spelling, wrong word and word order. The thesis also deals with the remedial treatment which may hopefully reduce the number of possible mistakes. The aim of the thesis is to compare the level of students' English language knowledge at the Secondary School of Tourism in Prague where the author works as an English teacher.

Wind energy has been growing rapidly during recent years. This paper aims to estimate the impact of wind power generation on the Swedish wholesale electricity price, using monthly time series data over the periods 2000-2016. The error-correction model is used to measure the price effect by including other factors that influence the electricity supply and demand. Thefindings suggest that the impact of changes in wind power production on the wholesale priceof electricity is negative in the short-term. When the wind power production increases by 1%, the wholesale electricity price decreases with 0.08%. Furthermore, the magnitude of the coefficient increases to 0.10% in the long-term.

The Groningen Meaning Bank (GMB) project develops a corpus with rich syntactic and semantic annotations. Annotations in GMB are generated semi-automatically and stem from two sources: (i) Initial annotations from a set of standard NLP tools, (ii) Corrections/refinements by human annotators. For example, on the part-of-speech level of annotation there are currently 18,000 of those corrections, so called Bits of Wisdom (BOWs). For applying this information to boost the NLP processing we experiment how to use the BOWs in retraining the part-of-speech tagger and found that it can be improved to correct up to 70% of identified errors within held-out data. Moreover an improved tagger helps to raise the performance of the parser. Preferring sentences with a high rate of verified tags in retraining has proven to be the most reliable way. With a simulated active learning experiment using Query-by-Uncertainty (QBU) and Query-by- Committee (QBC) we proved that selectively sampling sentences for retraining yields better results with less data needed than random selection. In an additional pilot study we found that a standard maximum-entropy part-of-speech tagger can be augmented so that it uses already known tags to enhance its tagging decisions on an entire sequence without retraining a new model first. Powered by...

Cast aluminum alloys are common in automotive and aerospace applications due to their high strength-to-density ratio. Fracture data for cast aluminum alloys, such as fatigue life, tensile strength and elongation, are heavily affected by the structural defects, such as pores and bifilms. There have been numerous studies in which either fatigue performance or tensile deformation were characterized and linked to casting defects. However, a comprehensive study that correlates tensile and fatigue properties has not been reported. The present study is motivated to fill this gap. The main objective of the investigation is to analyze the link between tensile and fatigue performance of commonly used cast aluminum alloys, and determine whether fatigue performance of cast aluminum alloys can be predicted. To accomplish this task, four research questions were developed: (i) how well do equations developed to account for mean stress effects perform in cast aluminum alloys, especially in datasets with various levels of structural quality, (ii) is the strong correlation between fatigue life and structural quality index obtained from tensile data reported for A206 alloy castings applicable to other aerospace and automotive casting alloys, (iii) how do methods to estimate high cycle fatigue from tensile data perform with aluminum castings, and (iv) can the axial fatigue performance of an A356-T6 casting be predicted from rotating beam fatigue data. Among the three mean stress correction models analyzed by using seven datasets from the literature, the one developed by Walker with an adjustable exponent has provided the best fit. It has been hypothesized that the adjustable Walker parameter is related to the structural quality index, QT, estimated from tensile data. Results have shown that there is indeed a strong correlation between QT and the Walker parameter. Moreover the parameters of the xvi Weibull distribution estimated from corrected data have been found to be strongly influenced by the mean stress correction method used. Tensile and fatigue life data for 319, D357 and B201 aluminum alloy castings reported in the literature have been reanalyzed by using a maximum likelihood method to estimate Basquin parameters in datasets with run-outs, Weibull statistics for censored data and mean stress correction. After converting tensile data to QT, a distinct relationship has been observed between the expected fatigue life and mean quality index for all alloys. Moreover, probability of survival in fatigue life has been found to be directly linked to the proportions of the quality index distributions in two different regions, providing further evidence about the strong relationship between elongation, i.e., structural quality, and fatigue performance [1]. Specimen geometry has been found to make the largest difference whereas the two aerospace alloys, B201 and D357, with distinctly different microstructures, have followed the same relationship, reinforcing the findings in the literature that fatigue life in aluminum castings is mainly determined by the size distribution and number density of structural defects. Six methods to predict fatigue life from tensile data have been compared by using data from the literature as well as the experimental A356 data developed in this study. Results have shown that none of the six methods provide reliable results. The consistently poor performance of the methods developed for steels and wrought alloys can be attributed to the major structural defects, namely bifilms, in aluminum castings. A new method to estimate the S-N curve from tensile data have been developed by using data for seventy-one S-N curves have been collected and Basquin parameters have been determined. Analysis showed that there is a strong relationship between QT and the Basquin exponent. xvii The Basquin parameters estimated by using the empirical relationships developed in the present study have provided better fits to the same datasets tested for the six methods. Hence the model developed in this study is proposed as the most reliable method to estimate high cycle fatigue properties. Finally, three methods to convert rotating bending fatigue test results to uniaxial fatigue data have been investigated by using the data developed in this study. Results have indicated that the method developed by Esin, in which both the fatigue life and alternating stress are corrected, provide the best estimate. Analyses of fracture surfaces of broken specimens via scanning electron microscopy have shown that tensile, axial fatigue and rotating beam fatigue properties are all strongly influenced by the same structural defects, confirming the validity of the approach taken in this study.

Master of Science / Department of Statistics / Susan J. Brown / Haiyan Wang / DNA hybridization microarray technologies have made it possible to gain an unbiased perspective of whole genome transcriptional activity on such a scale that is increasing more and more rapidly by the day. However, due to biologically irrelevant bias introduced by the experimental process and the machinery involved, correction methods are needed to restore the data to its true biologically meaningful state. Therefore, it is important that the algorithms developed to remove any sort of technical biases are accurate and robust. This report explores the concept of background correction in microarrays by using a real data set of five replicates of whole genome tiling arrays hybridized with genetic material from Tribolium castaneum. It reviews the literature surrounding such correction techniques and explores some of the more traditional methods through implementation on the data set. Finally, it introduces an alternative approach, implements it, and compares it to the traditional approaches for the correction of such errors.

This research analyzes the operating performance of two commercially available passive infrared counters (PICs) of pedestrians as a function of site, summer, fall and winter seasons in terms of counter sensitivity. Three sites were selected for field analysis in Winnipeg, Canada. Based on a sample of 24,690 people counted by the two PICs from July 2014 to February 2015, this research found that with a 95 percent confidence, Eco-Counter’s sensitivity ranged from 73 to 97 percent while TRAFCO’s ranged from 57 to 97 percent related to people occlusion. On weekdays, Eco-Counter’s absolute error was 16 percent and TRAFCO’s was 18 percent. On weekends, Eco-Counter’s absolute error was 18 percent and TRAFCO’s was 21 percent. In addition to people occlusion, site, seasons, and time of week (weekday and weekend) were found to affect the operating performance of the PICs. Correction factors were also calculated per counter, site, and seasons. / May 2016

Magister Philosophiae - MPhil / Life expectancy in South African has been fluctuating following the global trends that affects both developed and developing countries. In South Africa the average life expectancy from 1994 to 1996 was higher with an average of 61,3 years. As from 1997 to 1999 it declined to an average of 58,4 years. The difference in years between 1994-1996 and 1997- 1999 was 2,9 years. From 2000-2002, life expectancy continued to decline to an average of 54,6 years. Life expectancy declined in a constant proportion from 2003-2005 and 2006-2008. In 2003-2005 it slightly declined to 52 years and in 2004-2007 it declined to 42,0 years. Life expectancy escalated after the mentioned years to 54,4 years between 2009-2011 and from 2012-2013 life expectancy was 54,0 years on average. This study examined factors or variables that verify the socioeconomic determinants of life expectancy in post-apartheid South Africa. Understanding the relationship between life expectancy and the socioeconomic variables was based on three objectives. The main objective for this study was to determine the impact of socioeconomic variables and health policy efforts on life expectancy, seeking an in-depth understanding by investigating the causality relationship between life expectancy and socioeconomic variables thus later investigating the difference between male and female’s life expectancy. This study was motivated by the fluctuating life expectancy in South Africa. The fluctuation in life expectancy were thus studied in relation to socioeconomic determinants which are government health expenditure, government education expenditure, GDP per capita, total fertility rate, urban population, access to sustainable drinking water and undernourishment. The mentioned variables were used as socioeconomic determinants of life expectancy during post-apartheid South Africa.

In the field of computer vision and more specifically multi-camera systems color correction is an important topic of discussion. The need for color-tone similarity among multiple images that are used to construct a single scene is self-evident. The strength and weaknesses of color- correction methods can be assessed by using metrics to measure structural and color-tone similarity and timing the methods. Color transfer has a better structural similarity than histogram-based prefiltering and a worse color-tone similarity. The color transfer method is faster than the histogram-based prefiltering. Color transfer is a better method if the focus is a structural similar image after correction, if better color-tone similarity at the cost of structural similarity is acceptable histogram-based prefiltering is a better choice. Color transfer is a faster method and is easier to run with a parallel computing approach then histogram-based prefiltering. Color transfer might therefore be a better pick for real-time applications. There is however more room to optimize an implementation of histogram-based prefiltering utilizing parallel computing.

Student Number : 9404757K - PhD thesis - School of Physiology - Faculty of Science / Natural secretions of oestrogen and progesterone vary according to menstrual phase in eumenorrhoeic women and have an affect on physiological systems that could consequentially influence exercise performance. In a series of menstrual phase comparative investigations (considering mainly the early follicular (EF), late follicular (LF) and mid-luteal (ML) phase), I aimed to elucidate the physiological relation of oestrogen and progesterone to fat metabolism and ventilation during endurance exercise and exercise performance. I measured plasma free fatty acid (FFA) kinetics during prolonged exercise from an intravenous infusion of K+[1-13C]palmitate in eumenorrhoeic women and found the change in palmitate rate of appearance and disappearance between menstrual phases to be related to the oestrogen/progesterone (E/P) ratio. Overall the results from this study suggest that oestrogen promotes increases in FFA availability during exercise, while progesterone mitigates this response. Therefore, a high oestrogen concentration and E/P ratio in the ML phase is required in order to produce a favourable FFA metabolic response for endurance events. In order to estimate plasma FFA oxidation rate using carbon-FFA tracers, the acetate correction factor which accounts for carbon-label retention in secondary metabolic pools, must be applied. When I derived the acetate correction factor by measuring fractional recovery of carbon-13 in expired CO2 from a constant infusion of Na+[1- 13C]acetate during submaximal exercise we found the acetate correction factor to be significantly lower in the ML than EF phase (average change from EF phase -1.8± 0.5%, p<0.05). Failure to account for the lower correction factor in the ML phase will result in a 6% underestimation of plasma FFA oxidation rate. Furthermore, since human serum albumin (HSA) routinely used as a carrier for parenteral delivery of FFA tracers has various disadvantages, I tested the feasibility of using 2-hydroxypropyl-b- cyclodextrin (HP-b-CD) as a possible alternative carrier. A comparison of expired CO2 enrichment following HSA-FFA (7.4±2.0 %o) and HP-b-CD-FFA (8.6±2.1%o) infusion during exercise showed that the HP-b-CD does not compromise natural in vivo behaviour of the FFA tracer (p=0.4). Progesterone-induced hyperventilation is occasionally reported during exercise in the luteal phase. I found that the change in ventilatory parameters (minute ventilation and respiratory rate) during exercise from EF to ML phase is related to both the oestrogen and progesterone concentration in the ML phase. However, the associated increase in respiratory rate throughout prolonged exercise in the ML versus EF phase did not increase metabolic demand and therefore could not be expected to exacerbate fatigue. When exercise performance was evaluated by means of a cycling time trial, I found a trend for best performance in the LF phase versus the EF phase (8 of 11 subjects improved by 5.2±2.9%, p=0.027), while no differences occurred between other menstrual phases. Metabolic and performance benefits of oestrogen may be concealed in the ML phase by the coincident increase in progesterone in this phase and thus a significant effect is often only evident with a high E/P ratio in the ML phase. The transient LF phase, characterised by the pre-ovulatory surge in oestrogen, reveals the maximum benefits of oestrogen on metabolism and performance during submaximal exercise.