|
|
Mining Parallel Corpora from the Web
Kúdela, Jakub ; Holubová, Irena (advisor)
Title: Mining Parallel Corpora from the Web Author: Bc. Jakub Kúdela Author's e-mail address: jakub.kudela@gmail.com Department: Department of Software Engineering Thesis supervisor: Doc. RNDr. Irena Holubová, Ph.D. Supervisor's e-mail address: holubova@ksi.mff.cuni.cz Thesis consultant: RNDr. Ondřej Bojar, Ph.D. Consultant's e-mail adress: bojar@ufal.mff.cuni.cz Abstract: Statistical machine translation (SMT) is one of the most popular ap- proaches to machine translation today. It uses statistical models whose parame- ters are derived from the analysis of a parallel corpus required for the training. The existence of a parallel corpus is the most important prerequisite for building an effective SMT system. Various properties of the corpus, such as its volume and quality, highly affect the results of the translation. The web can be considered as an ever-growing source of considerable amounts of parallel data to be mined and included in the training process, thus increasing the effectiveness of SMT systems. The first part of this thesis summarizes some of the popular methods for acquiring parallel corpora from the web. Most of these methods search for pairs of parallel web pages by looking for the similarity of their structures. How- ever, we believe there still exists a non-negligible amount of parallel...
|
|
|
Advanced Decomposition Methods in Stochastic Convex Optimization
Kůdela, Jakub ; Fabian, Csaba (referee) ; Šmíd,, Martin (referee) ; Popela, Pavel (advisor)
Při práci s úlohami stochastického programování se často setkáváme s optimalizačními problémy, které jsou příliš rozsáhlé na to, aby byly zpracovány pomocí rutinních metod matematického programování. Nicméně, v některých případech mají tyto problémy vhodnou strukturu, umožňující použití specializovaných dekompozičních metod, které lze použít při řešení rozsáhlých optimalizačních problémů. Tato práce se zabývá dvěma třídami úloh stochastického programování, které mají speciální strukturu, a to dvoustupňovými stochastickými úlohami a úlohami s pravděpodobnostním omezením, a pokročilými dekompozičními metodami, které lze použít k řešení problému v těchto dvou třídách. V práci popisujeme novou metodu pro tvorbu “warm-start” řezů pro metodu zvanou “Generalized Benders Decomposition”, která se používá při řešení dvoustupňových stochastických problémů. Pro třídu úloh s pravděpodobnostním omezením zde uvádíme originální dekompoziční metodu, kterou jsme nazvali “Pool & Discard algoritmus”. Užitečnost popsaných dekompozičních metod je ukázána na několika příkladech a inženýrských aplikacích.
|
|
|
Mining Parallel Corpora from the Web
Kúdela, Jakub ; Holubová, Irena (advisor) ; Helcl, Jindřich (referee)
Title: Mining Parallel Corpora from the Web Author: Bc. Jakub Kúdela Author's e-mail address: jakub.kudela@gmail.com Department: Department of Software Engineering Thesis supervisor: Doc. RNDr. Irena Holubová, Ph.D. Supervisor's e-mail address: holubova@ksi.mff.cuni.cz Thesis consultant: RNDr. Ondřej Bojar, Ph.D. Consultant's e-mail adress: bojar@ufal.mff.cuni.cz Abstract: Statistical machine translation (SMT) is one of the most popular ap- proaches to machine translation today. It uses statistical models whose parame- ters are derived from the analysis of a parallel corpus required for the training. The existence of a parallel corpus is the most important prerequisite for building an effective SMT system. Various properties of the corpus, such as its volume and quality, highly affect the results of the translation. The web can be considered as an ever-growing source of considerable amounts of parallel data to be mined and included in the training process, thus increasing the effectiveness of SMT systems. The first part of this thesis summarizes some of the popular methods for acquiring parallel corpora from the web. Most of these methods search for pairs of parallel web pages by looking for the similarity of their structures. How- ever, we believe there still exists a non-negligible amount of parallel...
|
|
|
Time minimization for vehicles passing a given trajectory
Suja, Jerguš ; Kůdela, Jakub (referee) ; Porteš, Petr (advisor)
This bachelor’s thesis deals with model building of automobile movement and it’s optimal speed in every single point of trajectory, whit aim on minimal lap time. It uses a real data, which was measured on Masaryk circuit in Brno. Contents of thesis are analysing effecting forces on vehicle’s dynamic and then building mathematic model. The main purpose is finding optimal speed characterization and compare with real data. Then, we modify model with adjustable wing and present advantage of this customization. Simulation was done by program MATLAB.
|
|
|
Algorithms for Detection and Correction of Local Degradations in Digital Audio Signals
Kúdela, Jakub ; Toropila, Daniel (advisor) ; Petříček, Martin (referee)
Title: Algorithms for Detection and Correction of Local Degradations in Digital Audio Signals Author: Jakub K'udela Author's e-mail address: jakub.kudela@gmail.com Department: Department of Theoretical Computer Science and Mathematical Logic Thesis Supervisor: Mgr. Daniel Toropila Supervisor's e-mail address: daniel.toropila@mff.cuni.cz Abstract: Local degradations in audio signal are discontinuities in their wave- forms. They are caused by the nature of the recording process, or by aging of or damage to the recording medium. In many cases these discontinuities are un- wanted while listening, and so there exists a number of methods, whose aim is to restore degraded recordings. In the introduction, this thesis informs the reader about selected algorithms for detection and correction of local degradations in digital audio signals. One of the discussed algorithms is a custom aplication of artificial neural networks to the given problem. The implementation of selected algorithms and experiments are both parts of the thesis. The goal of the exper- iments is to both objectively and subjectively compare the performances of the selected algorithms. The thesis proposes a method for the objective evaluation of the quality of detection and correction, which, as will be shown, largely cor- responds to the subjective...
|
|
|
Mining Parallel Corpora from the Web
Kúdela, Jakub ; Holubová, Irena (advisor)
Title: Mining Parallel Corpora from the Web Author: Bc. Jakub Kúdela Author's e-mail address: jakub.kudela@gmail.com Department: Department of Software Engineering Thesis supervisor: Doc. RNDr. Irena Holubová, Ph.D. Supervisor's e-mail address: holubova@ksi.mff.cuni.cz Thesis consultant: RNDr. Ondřej Bojar, Ph.D. Consultant's e-mail adress: bojar@ufal.mff.cuni.cz Abstract: Statistical machine translation (SMT) is one of the most popular ap- proaches to machine translation today. It uses statistical models whose parame- ters are derived from the analysis of a parallel corpus required for the training. The existence of a parallel corpus is the most important prerequisite for building an effective SMT system. Various properties of the corpus, such as its volume and quality, highly affect the results of the translation. The web can be considered as an ever-growing source of considerable amounts of parallel data to be mined and included in the training process, thus increasing the effectiveness of SMT systems. The first part of this thesis summarizes some of the popular methods for acquiring parallel corpora from the web. Most of these methods search for pairs of parallel web pages by looking for the similarity of their structures. How- ever, we believe there still exists a non-negligible amount of parallel...
|
|
|
Models for network flow identification
Smejkalová, Veronika ; Kůdela, Jakub (referee) ; Pavlas, Martin (advisor)
The bachelor thesis is focused on development of a mathematical model of network flow in waste managment. The theoretical part of thesis deals with graph theory, network flow and optimization. The next chapter introduces the mathematical model. The model was compiled with the use of Information systems of waste management. The data are very often burdened by error. The task of the model is revelation of those errors to make the data useful. This thesis also includes suggestions for further development.
|
|
|
Stochastic optimization in AIMMS
Kůdela, Jakub ; Mrázková, Eva (referee) ; Popela, Pavel (advisor)
Tato diplomová práce uvádí základní poznatky matematického a především stochastického programování. Navíc se zabývá použitím softwaru AIMMS při vytváření a řešení optimalizačních problémů. Naším hlavním cílem je naprogramovat v softwaru AIMMS několik metod řešení problémů stochastického programování a ukázat jejich použití a užitečnost na vybraných problémech. Jedním z problémů, který jsme si zvolili, je model spalovny. Všechny AIMMS programy, které v našem textu použijeme a popíšeme, a jejich zdrojové kódy budou přiloženy v dodatcích.
|
|
|
Optimization problems in AIMMS
Kůdela, Jakub ; Popela, Pavel (referee) ; Mrázková, Eva (advisor)
This bachelor's thesis deals with the usage of system AIMMS for solving and creating optimization models. After the introduction to theoretical background of optimization problems there is an example of modeling a selected problem and its implementation to AIMMS. Each of the selected problems has its graphical user interface where an end user can simply modify the data of the model. All the source codes of programs described above are attached.
|
|
|
Cloud solutions and their deployment
Kúdela, Jakub ; Matuštík, Ondřej (advisor) ; Šperková, Lucie (referee)
Bachelor thesis "Cloud solutions and their deployment" introduces topic of cloud computing as a significant trend to provide services over the Internet without using internal infrastructure or other resources. The first part of thesis defines term of cloud computing and introduces basic principles. As a phenomenon of today's IT world, cloud can offer many advantages to businesses and provide them with significant competitive advantage. Even in this new technology, however, exist obstacles and some challenges may occur that businesses need to take into consideration. It is possible to distinguish different service models and deployment models that are presented in the thesis.. The aim of this thesis is to present a comprehensive look at the topic of cloud computing and subsequently provide a practical, real-world use case of deploying cloud on a company. In the practical part of thesis is presented a company with its existing architecture, which is interested in leveraging cloud solutions. The paper describes the core business of the company, its goals, the current state of architecture and requirements placed on the transition. There are some questions raised in the company such as selection of a vendor, on-premise vs. cloud solution, security and selecting the right service models. End of the thesis is devoted to a brief presentation of concrete solutions that ensure a transition to the cloud in the company.
|
guest ::
RSS feed.