|
|
STATISTICAL LANGUAGE MODELS BASED ON NEURAL NETWORKS
Mikolov, Tomáš ; Zweig, Geoffrey (referee) ; Hajič,, Jan (referee) ; Černocký, Jan (advisor)
Statistické jazykové modely jsou důležitou součástí mnoha úspěšných aplikací, mezi něž patří například automatické rozpoznávání řeči a strojový překlad (příkladem je známá aplikace Google Translate). Tradiční techniky pro odhad těchto modelů jsou založeny na tzv. N-gramech. Navzdory známým nedostatkům těchto technik a obrovskému úsilí výzkumných skupin napříč mnoha oblastmi (rozpoznávání řeči, automatický překlad, neuroscience, umělá inteligence, zpracování přirozeného jazyka, komprese dat, psychologie atd.), N-gramy v podstatě zůstaly nejúspěšnější technikou. Cílem této práce je prezentace několika architektur jazykových modelůzaložených na neuronových sítích. Ačkoliv jsou tyto modely výpočetně náročnější než N-gramové modely, s technikami vyvinutými v této práci je možné jejich efektivní použití v reálných aplikacích. Dosažené snížení počtu chyb při rozpoznávání řeči oproti nejlepším N-gramovým modelům dosahuje 20%. Model založený na rekurentní neurovové síti dosahuje nejlepších publikovaných výsledků na velmi známé datové sadě (Penn Treebank).
|
|
|
Implementation of Statistical Compression Methods
Beněk, Tomáš ; Polok, Lukáš (referee) ; Bařina, David (advisor)
This thesis deals with the statistical compression methods. This work describes basic theoretical introduction to the data compression and the most important representatives of statistical methods. There is also described compression process based on the Burrows-Wheeler algorithm, which is implemented in the form of portable library in C++ programming language. This work compares effectiveness of each method on suitable data corpus, which contains various data types and according to achieved results discusses suitability of usage of methods and possible improvements.
|
|
|
Compression Methods Based on Context Modelling
Kozák, Filip ; Drábek, Vladimír (referee) ; Bařina, David (advisor)
Purpose of this thesis is to describe the context-based compression methods and their application to multimedia data. There is described the principle of arithmetic coding and prediction by partial matching method, including creation of the probability model. There are also described multimedia data and the basic principles of their compression. The next section presents compression methods, that I implemented at work and their results.
|
|
|
Application of Slepian sequences for Spectral Analysis
Šimeček, Vít ; Jirgl, Miroslav (referee) ; Mihálik, Ondrej (advisor)
This thesis is focused on the Digital signal processing using the Slepian transform, which is a option of gaining a signal spectrum. The signal is transformed to the Slepian orthogonal base, which can be generated in Matlab by the function "DPSS". Furthermore, this spectrum is in the thesis compared with classic Fourier spectrum and cosine spectrum. Further, a data compression is computed and bases are compared to each other. By the Slepian base, a simple signal is extrapolated.
|
|
|
Using Cellular Automata for Data Compression
Polák, Marek ; Trunda, Otakar (advisor) ; Mráz, František (referee)
In this thesis we research the possibilities of using cellular automata for lossless data compression. We describe the classification of cellular automata and their current usage. We study the properties of various types of elementary cellular automata (i.e. Wolfram rules), describe their equivalence classes, the ways of forward as well as backward simulation, we examine the rules with interesting behavior. The states provided by these rules are evaluated in terms of their orderliness (e.g. the ratio of living cells or approximation of entropy). We implement some standard compression algorithms and compare them in terms of usability for best rated states. By application of acquired knowledge we propose a new compression algorithm, test it on text and image data and compare the results with traditional compression algorithms. Powered by TCPDF (www.tcpdf.org)
|
|
|
Sinumerik Operate tools for programming of CNC machines
Krejčík, Jan ; Sliwková, Petra (referee) ; Polzer, Aleš (advisor)
The diploma thesis is focused on possibilities of spline approximation and interpolation of CAD imported by CAD Reader to control system. First chapter describes mathematical meaning of curves, the second chapter shows possibilities of approximation and interpolation in control system Sinumerik 840D and the third chapter applies these possibilities on practical example.
|
|
|
Implementation of Statistical Compression Methods
Štys, Jiří ; Polok, Lukáš (referee) ; Bařina, David (advisor)
This thesis describes Burrow-Wheeler compression algorithm. It focuses on each part of Burrow-Wheeler algorithm, most of all on and entropic coders. In section are described methods like move to front, inverse frequences, interval coding, etc. Among the described entropy coders are Huffman, arithmetic and Rice-Golomg coders. In conclusion there is testing of described methods of global structure transformation and entropic coders. Best combinations are compared with the most common compress algorithm.
|
|
|
Analysis of JPEG 2000 Settings
Vrtělová, Lucie ; Klíma, Ondřej (referee) ; Bařina, David (advisor)
This thesis is focused on analysis of JPEG 2000 settings. The aim of this thesis is to explore and compare each settings and different implementation. The analysis was done with two libraries, which implements JPEG 2000 - Kakadu and OpenJPEG. Data was processed by framework written in Python3. Created solution provides comprehensive view on using each settings of JPEG 2000 on specific data. Results of this thesis allow faster using of JPEG 2000 on specific data.
|
|
|
Using Cellular Automata for Data Compression
Polák, Marek ; Trunda, Otakar (advisor) ; Mráz, František (referee)
In this thesis we research the possibilities of using cellular automata for lossless data compression. We describe the classification of cellular automata and their current usage. We study the properties of various types of elementary cellular automata (i.e. Wolfram rules), describe their equivalence classes, the ways of forward as well as backward simulation, we examine the rules with interesting behavior. The states provided by these rules are evaluated in terms of their orderliness (e.g. the ratio of living cells or approximation of entropy). We implement some standard compression algorithms and compare them in terms of usability for best rated states. By application of acquired knowledge we propose a new compression algorithm, test it on text and image data and compare the results with traditional compression algorithms. Powered by TCPDF (www.tcpdf.org)
|
|
|
Improvements of multistream compression
Unger, Lukáš ; Žemlička, Michal (advisor) ; Pokorný, Jaroslav (referee)
Multistream compression is based on a transformation significantly different from the ones commonly used for data compression. This Master thesis concerns with the use of said method for the compression of text files written in natural language. The main goal of the thesis is to find suitable preprocessing methods for text transformation, which would enable the Multistream compression to achieve better compression ratios, together with searching for the best methods for coding of individual streams. The practical part of the thesis deals with the implementation of several transformation algorithms into the XBW project.
|
guest ::
