Este trabalho avaliou, utilizando o Método dos Elementos Finitos, o comportamento mecânico em seis ensaios, quatro de flexo compressão com variação apenas no comprimento da haste de fixação, um de flexão ântero-posterior e um de flexão lateral, em um modelo tridimensional de fixador externo construído com auxílio do computador. Para a validação desses ensaios os mesmos testes também foram realizados na máquina universal de ensaio e depois comparados com o método dos elementos finitos. A rigidez no ensaio de flexão lateral obtida no método dos elementos finitos apresentou um valor 106% maior do que o obtido na máquina universal de ensaio, possivelmente por ocorrência de afrouxamento entre as partes adjacentes elo fixador externo no ensaio mecânico, e no ensaio de flexo compressão (número 1) a rigidez obtida pelo método dos elementos finitos foi 7,39% menor elo que a obtida na máquina universal de ensaio. No ensaio número 4 (flexo-compressão) os deslocamentos obtidos, correspondentes a aplicação de uma carga de 15,4 N, utilizando tanto o método dos elementos finitos como a máquina universal ele ensaio, divergiram em 8,5%. Já no ensaio número 5 (flexão ântero-posterior) a divergência foi de 0,9%, para os deslocamentos correspondentes a aplicação de uma carga de 14,4 N. O Métodos dos Elementos Finitos é um eficiente método para prever comportamento mecânico de estruturas complexas. / This work evaluates the mechanical behavior of an external device on a CAD (Computer Aided Design) 3D model in six tests, four of them of flexion-compression with a single variation of the external device stick length, one of longitudinal flexion and one of lateral flexion. They all were carried out applying the Method of the Finite Elements (MFE). For the comparison with the MFE the same mechanical testes were made. The stiffness achieved in the MFE showed a value 106% greater than that of the mechanical test. The test number one of flexion-compression showed a value 7,39% smaller in the MFE than that of the mechanical test. In the test number four (flexion-compression) the corresponding displacement after the application of a 15,4 N working stress showed a difference of 8,5%. The variation of the number five test was of 0,9% for the displacement after the application of a 14,4N working stress. The MEF showed efficient to evaluate the behavior of complex mechanical structures.

abstract: There have been multiple attempts of coupling neural networks with external memory components for sequence learning problems. Such architectures have demonstrated success in algorithmic, sequence transduction, question-answering and reinforcement learning tasks. Most notable of these attempts is the Neural Turing Machine (NTM), which is an implementation of the Turing Machine with a neural network controller that interacts with a continuous memory. Although the architecture is Turing complete and hence, universally computational, it has seen limited success with complex real-world tasks. In this thesis, I introduce an extension of the Neural Turing Machine, the Neural Harvard Machine, that implements a fully differentiable Harvard Machine framework with a feed-forward neural network controller. Unlike the NTM, it has two different memories - a read-only program memory and a read-write data memory. A sufficiently complex task is divided into smaller, simpler sub-tasks and the program memory stores parameters of pre-trained networks trained on these sub-tasks. The controller reads inputs from an input-tape, uses the data memory to store valuable signals and writes correct symbols to an output tape. The output symbols are a function of the outputs of each sub-network and the state of the data memory. Hence, the controller learns to load the weights of the appropriate program network to generate output symbols. A wide range of experiments demonstrate that the Harvard Machine framework learns faster and performs better than the NTM and RNNs like LSTM, as the complexity of tasks increases. / Dissertation/Thesis / Masters Thesis Computer Science 2020

This thesis is about How External Uncertainties Affect the Innovation Process. Written during the spring term of 2020 by Simon Algotsson and Johan Öhlander. The thesis main goal is to generate knowledge about the properties and sources of external uncertainties and create an understanding of how they can come to affect an innovation process. This research encourages organizations that are planning to participate, or currently resides in an innovation process to give it a read. Anyone who seeks a deeper understanding of the impact of external uncertainties may use our findings as a source of inspiration. The research question we have answered is: How do external uncertainties affect the innovation process? As the title and research question shows, innovation and uncertainties are the two most common denominators in this work. Presented in the theoretical framework is previous research done concerning the innovation process, and what it consists of. As well as how other researchers describe different types of uncertainties. We have also constructed our own model of how external uncertainties can give rise to internal uncertainties. For this type of research, a qualitative method has been selected, since it enabled us to go in-depth in one specific innovation process. We have conducted two interviews with the CEO and CFO of a company referred to as Gamma. They have both shared their own perspectives of the innovation process their company has gone through. The data collected from the interview has been transcribed separately and is later presented in the empirical evidence. The final sections of this thesis include the analysis and the conclusion. In these chapters we draw parallels between the research presented in the theoretical framework and Gamma's innovation process. The process we have investigated for this thesis encountered several uncertainties, both internal and external. In the analysis we present the authors own model of how external uncertainties came to affect Gamma's innovation process. The conclusion discusses the significant findings of the research such as how Gamma's innovation turned into a 'black hole' for the profits generated by the company.

The goal of this research was to study the polarization effect on the achievability of mode locked laser signal in accordance to polarization dependency. Polarization characterization was studied in prior to mode locking to determine semiconductor optical amplifiers (SOA) polarization dependency. It was found that per SOA manufacturer a variation of dependency was recognized. Two setups were suggested based on dissipative soliton resonance (DSR) schemes, the external DSR with regenerative feedback and the total internal reflection (TIR) DSR. The laser operated at 1530nm for multiple SOA manufacturers and was capable of generating ultrashort pulses with different durations as listed in Table (3.2.3) and Table (3.3.2). For the external mode-lock, the maximum achieved average output power of 3.40mW corresponding to ~77.15fJ of output pulse energy through continuous pumping by gain modulation of an SOA used as the active medium. While the TIR mode-lock the maximum achieved average output power for mode-locked signal of 5.25mW corresponding to ~23.73fJ of output pulse energy.It was concluded that the output pulse of the external DSR setup showed polarization dependency for two SOAs A and C and provided mode locked signals with pulse width ranging between 1200ps and 1000ps with repetition rate approximately of 35MHz and 101MHz respectively. The TIR DSR had a variation in results also to be polarization related. While SOA D did not provide mode-locked signal output. The least polarization dependent SOAs E and X did not provide distinctive stable mode-locked signal. While increased sensitivity towards polarization however resulted in stronger mode locked signal for SOAs A and C with FWHM between 200ps and 350ps and a repetition rate of approximately 20MHz. For future work, it is recommended to adjust the setup length in addition to fine tuning polarizer and filters to improve pulse stability and eliminate the noise effect.

京都大学 / 0048 / 新制・課程博士 / 博士(農学) / 甲第22075号 / 農博第2367号 / 新制||農||1072(附属図書館) / 学位論文||R1||N5229(農学部図書室) / 京都大学大学院農学研究科地域環境科学専攻 / (主査)教授 近藤 直, 准教授 小川 雄一, 教授 清水 浩 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

京都大学 / 0048 / 新制・課程博士 / 博士(社会健康医学) / 甲第22835号 / 社医博第111号 / 新制||社医||11(附属図書館) / 京都大学大学院医学研究科社会健康医学系専攻 / (主査)教授 佐藤 俊哉, 教授 川上 浩司, 教授 木村 剛 / 学位規則第4条第1項該当 / Doctor of Public Health / Kyoto University / DFAM

The purpose of the master thesis is a proposal of an effective personnel marketing system in a company operating in a field of information technology resulting in meeting needs of potential and current employees. The thesis includes an analysis of external and internal personnel marketing supported by quantitative research among information technology programme students of Brno universities and current employees. As emerged from the analysis, shortcomings can be found in a company promotion via social media and web page as well as in a professional development of employees. Based on the acquired results and theoretical overview particular proposals of effective personnel marketing have been made.

The diploma thesis is focused on personnel marketing and its application in the manufacturing company. It characterizes the concept, principles and significance of the HR marketing. It also contains an analysis of the current company’s situation which is supported by conducted qualitative research. According to this analysis there are formulated suggestions that can contribute to the maintenance of the necessary human resources.

This master thesis focuses on analyzing the marketing management company Diplomat Dental Ltd. The aim of this work is to analyze the application of marketing management in the company and considering the facts suggest solutions that will lead to an increase in marketing activities and the total company running. The company is analysed from the internal and external aspect through marketing analyses as marketing mix, value chain analysis, SWOT analysis, SLEPT analysis and others.

In Life Sciences Education, the use of educational external representations (ERs) such as diagrams, models and animations are increasingly appearing in learning and teaching resources. However, their effectiveness is limited if learners experience learning difficulties due to lack of visual literacy and spatial ability skills to work with ERs. The study explored the level of visual literacy of 225 Grade 10 Life Sciences learners in cytology across six secondary schools in Pretoria, Gauteng. It was theorised that ERs need to be integrated in the Life Sciences curricula to develop learners’ visual literacy and spatial ability skills. With this theory, the study aimed to explore the visual literacy of Grade 10 Life Sciences learners and the influence of gender and school location on the visual literacy and spatial ability skills of the learners. Through a quantitative research method a Life Sciences visual literacy questionnaire and a spatial ability test were used to collect data. Collected data was analysed descriptively and inferentially through Statistical Package Social Sciences Version 23. The results showed that most Grade 10 Life Sciences learners lack average visual literacy skills. Furthermore, the results showed that gender doesn’t play a role on the learners’ performance in visual literacy skills as both genders performed equally in both tests, On the other hand, the results showed that the location of the school (urban, rural or township) has an effect on the learners’ performance in visual literacy skills. Teachers need to incorporate different ERs that would stimulate different senses and which will also enhance learners’ visual literacy and spatial ability skills in their lessons. A conclusion and some recommendations for future research are given. / Dissertation (MEd)--University of Pretoria, 2017. / Science, Mathematics and Technology Education / MEd / Unrestricted