近年來資料產生的數量遠超過過去可處理的數量，以現今的個人電腦使用傳統的方法已經無法處理大資料的運算與分析，所以改善傳統的方法與平行化為必經的方向，本論文以拆解合成-多元尺度法的平行化為主要討論對象，除了介紹Python程式語言及其相關套件如何撰寫平行化程式，我們將拆解合成-多元尺度法從原本的單核心版本改進為多核心版本，並且探索拆解合成-多元尺度法在平行化過程中的計算效能，藉以了解拆解合成-多元尺度法在平行化計算時的參數要如何設定，使得平行化的SC-MDS可以有最高的計算效率。經實驗證明多核心底下的SC-MDS平行化又把SC-MDS單核心的效能做個再次的提升。 / In recent years, the number of generated data is growing fast such that it is infeasible to process by using traditional methods. So improving traditional methods and developing paralled computing methods are important issues. The main contribution of this thesis is to delelope the parallel version of the split-and-combine multidimensional scaling method(SC-MDS). We will fistly introduce fundamental python program, the basic python packages and the python multi-core program. Then we will implement the serial core version of SC-MDS to the multi-core version. Moreover, we will discover the efficiency of the multi-core version of SC-MDS. Then we can understand how to determine the parameters of the parllel version of SC-MDS. By our experimental results, we successfully implement the serial core of SC-MDS to the faster parallel version of SC-MDS.

Thesis (M.S. in computer science)--Washington State University, May 2009. / Title from PDF title page (viewed on May 26, 2009). "School of Electrical Engineering and Computer Science." Includes bibliographical references (p. 41-44).

This work is focused on classifying photos which are uploaded on dating service Lidé.cz. Pictures are classified into two categories based on whether they contain pornographic content or not. Convolutional neural networks are used for classification and these neural networks are taught by using Caffe framework. The results of this work fulfilled all requirements from Seznam.cz, a.s. company. Classification accuracy of the best model on created testing dataset with 5643 photos was 93,64 % and the time for classification of photography is low enough to perform classification in real time. The first part contains an analysis of the current approaches for image classification. The second part focuses on the analysis and draft of the solution and the third part describes the implementation of the solution and the testing of neural networks models.

The purpose of this diploma thesis is to introduce a basic summary of landscape connectivity calculations mainly focusing on the Conefor program and the effect of input parameters on output landscape connectivity indices. One of the main goals is to create an ArcMap Toolbox to simplify the calculation of landscape connectivity indices for many areas of interest. Basic squares that cover the whole Czech Republic were chosen as main areas of interest. These units are used for example for mapping the abundance of flora and fauna. The same indices were calculated for NUTS3 and LAU1 areas in order to compare effect of size of input areas. This thesis provides a simple solution in the form of a Python Toolbox for Esri ArcMap to calculate overall indices of landscape connectivity through a Conefor command line application for a large number of input areas. Forest connectivity for different dispersal distances and input areas of interest was calculated with this Toolbox. As an input vector layer of forests, CORINE land cover data from mapping period 1990, 2000, 2006 and 2012 were used. Input dispersal distances were chosen in the range from 50 to 50.000 meters in order to comprehensively describe their effects on indices. Resultant forest and connectivity changes are represented by cartograms; the effect of input data on output indices is shown in graphs.

Submitted by João Arthur Martins (joao.arthur@ufpe.br) on 2015-03-12T17:26:20Z No. of bitstreams: 2 Dissertacao Thiago Varjao.pdf: 1253062 bytes, checksum: 2f617e0f9c6d1b7386384337d729200a (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Approved for entry into archive by Daniella Sodre (daniella.sodre@ufpe.br) on 2015-03-13T13:24:30Z (GMT) No. of bitstreams: 2 Dissertacao Thiago Varjao.pdf: 1253062 bytes, checksum: 2f617e0f9c6d1b7386384337d729200a (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Made available in DSpace on 2015-03-13T13:24:30Z (GMT). No. of bitstreams: 2 Dissertacao Thiago Varjao.pdf: 1253062 bytes, checksum: 2f617e0f9c6d1b7386384337d729200a (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Previous issue date: 2013-08-30 / No mundo do desenvolvimento de software, Java representa um marco. Concebida na década de 90, alcançou enorme popularidade desde o início de sua utilização. Apresentada inicialmente como uma linguagem que trabalhava em sites para internet, o que não era possível na época, seu amadurecimento levou a criação de muitas implementações, permitindo-lhe, hoje, ser encontrada, além das páginas da internet, em desktops, celulares e diversos outros dispositivos, provendo grande segurança, sendo utilizada em grandes bancos e empresas que necessitam de estabilidade e portabilidade, para trafegar grande quantidade de dados. Apesar de toda influência da linguagem Java, baseada em seu poder e aplicabilidade, ela não costuma ser usada na computação científica por ter os típicos problemas computacionais acentuados pela forte tipagem de seus tipos primitivos, comprometendo sua aplicação na matemática computacional, mesmo em experimentos feitos, na implementação de bibliotecas intervalares que usam esses tipos primitivos. Este trabalho apresenta a extensão de uma biblioteca, fundamentada na matemática intervalar e aritmética de exatidão máxima, na linguagem Java. Além das funções potência, raiz quadrada, exponencial, logarítmica e trigonométricas, a nova versão da biblioteca foi incrementada com cálculo de probabilidades para as variáveis aleatórias Uniforme, Exponencial e Pareto. Por fim, foi desenvolvida uma proposta de benchmark para comparação de tecnologias aplicadas à matemática intervalar no qual foram confrontados os desempenhos da extensão aqui proposta com uma biblioteca intervalar em Python.

Denna studie handlar om en implementering av standardisering och automatisering i mindre organisationer. Jag skapade ett skript i Python som upptäcker uppladdade kalkylblad i en mapp som i sin tur var kopplad till en molnbaserad lösning. När skriptet upptäcker kalkylbladen samlar den nödvändig information och importerar en mall av en faktura. Skriptet fyller efterhand mallen med tidigare samlad information och sparar det i en separat mapp. Det originella kalkylbladet arkiveras för framtida referenser. Resultatet av en bättre, standardiserad arbetsrutin och automatiserat fakturasystem resulterade i en tidsbesparing på över 60% och i samband med det nya systemet vi såg en minskning på 75% av incidenter orsakade av dokumentslarv och miss kommunikation.

abstract: Neuron models that behave like their biological counterparts are essential for computational neuroscience.Reduced neuron models, which abstract away biological mechanisms in the interest of speed and interpretability, have received much attention due to their utility in large scale simulations of the brain, but little care has been taken to ensure that these models exhibit behaviors that closely resemble real neurons. In order to improve the verisimilitude of these reduced neuron models, I developed an optimizer that uses genetic algorithms to align model behaviors with those observed in experiments. I verified that this optimizer was able to recover model parameters given only observed physiological data; however, I also found that reduced models nonetheless had limited ability to reproduce all observed behaviors, and that this varied by cell type and desired behavior. These challenges can partly be surmounted by carefully designing the set of physiological features that guide the optimization. In summary, we found evidence that reduced neuron model optimization had the potential to produce reduced neuron models for only a limited range of neuron types. / Dissertation/Thesis / Doctoral Dissertation Neuroscience 2020

This diploma thesis describes a design and implementation of a system for monitoring events in passive optical networks, specifically in GPON networks. The main technologies used in the implementation of this system are Apache Kafka, Docker and the Python programming language. Within the created application, several filters are implemented. This filters obtain essential information from the captured frames in terms of traffic analysis on the given network. The result of the thesis is a functional system that from the captured GPON frames obtains information about the network traffic and stores them in the Apache Kafka, where the stored data is accessible for further processing. The work also provides examples of how to process the stored data, along with information about their meaning and structure.

The subject of this thesis is digitalization of bus transportation. The input is represented by a sequence of GNSS data which are transformed to the OpenStreetMap format. Doing so, it is enriched by the information OpenStreetMap format provides and it gains its positional advantages as well. Then this thesis deals with ways by which one can detect bus lines from this general sequence of GNSS coordinates. A bus line is recognized as a repeating trajectory, which satisfies criteria derived from its expectable or defined characteristics. A few clustering solutions are proposed and tested for their performance. On the basis of this testing, there is one solution chosen as the best performing one, to be the proposed solution of this thesis. The overall output will therefore be formed by automatic mapping of bus lines with no theoretical area limit and with minimum manual intervention needed. It lays the foundations for various intelligent real-time processes to be implemented as well as allowing for infrastructure to be processed for the statistics purposes or urban planning.

The topic of this thesis is registration of image sequences captured by experimental ophthalmoscope. It contains anatomical description of the visual system as well as the description of functions of selected ophthalmoscopic devices. The next covered topic is theoretical summary of registration process, which is followed by an overview of the used methods, which forms the basis of the design and implementation of the registration algorithm in the Python programming language. After implementation, the accuracy and computational complexity of a registration was evaluated. Tests of optimalization of the proposed approach were performed with regards to the obtained results, through which sufficiently accurate registration has been achieved, evaluated on the basis of Euclidean distances, standard deviation and visual observation. In case of high-quality recorded sequences, values of Euclidean distances ranged from 0.60 to 4.07 pixels on the contrary, values higher than 20 pixels occurred in the case of poor-quality recordings. Standard deviation values in recordings with high enough resolution have not reached worse results than 4.12.