|
|
Data and their clustering
Pilmann, Jindřich ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
This master thesis descripes known methods of data clustering and examines their possible application on data from the area of social networks. Because of this we recapitulated how we describes objects using data and which technics we use for specifying their similarity. After that we recapitulated known clustering methods and possibilities of their validation. Consequently we have suggested method how perform clustering in the social networks and we tested it. We have applied this method on data from the area of international trade in 2008. We have evaluated and summarized results of this experiments. In the end of this work we have suggested possibilities of further research in this area.
|
|
|
Processing of secondary structures in nucleic acids
Goldwein, Robert ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
This work explores and studies basic methods of bioinformatics - new and perspective branch of computer science. Introduces the term Bioinformatics, familiarizes with necessary biological background (DNA and RNA molecules, proteins, central dogma of molecular biology) and also with basic bioinformatics terms (biological sequence, primary and secondary structure). It also demonstrates the implementation of basic bioinformatics algorithms and their use with real data (on Foot-and-mouth disease virus) - motif finding, longest common subsequence and sequence alignment. This work also introduces higher structures of biological sequences, primarily with secondary structure of RNA molecule.
|
|
|
Doporučování zboží ve webových obchodech
Helmich, Jiří ; Sýkora, Ondřej (advisor) ; Kukačka, Marek (referee)
Electronic shopping recommendation systems are an integral part of most on-line stores. Nowadays, these systems are usually driven by Market Basket Analysis algorithms and by finding relations within the range of goods offered by the shop. However, for a recommendation system an action-based approach might be a viable alternative. The student will review existing approaches to recommendation based on reinforcement learning algorithms, and implement a prototype of a recommender system for a web store based on the results of the research.
|
|
|
RBF-networks with a dynamic architecture
Jakubík, Miroslav ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
In this master thesis I recapitulated several methods for data clustering. Two well known clustering algorithms, concretely K-means algorithm and Fuzzy C-means (FCM) algorithm, were described in the submitted work. I presented several methods, which could help estimate the optimal number of clusters. Further, I described Kohonen maps and two models of Kohonen's maps with dynamically changing structure, namely Kohonen map with growing grid and the model of growing neural gas. At last I described quite new model of radial basis function neural networks. I presented several learning algorithms for this model of neural networks, RAN, RANKEF, MRAN, EMRAN and GAP. In the end of this work I made some clustering experiments with real data. This data describes the international trade among states of the whole world.
|
|
|
Multi-layered neural networks and visualization of their structure
Drobný, Michal ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
The model of multi-layered neural networks of the back-propagation type is well-known for their universal approximation capability and even the standard back-propagation training algorithm used for their adjustment often provides results applicable to real-world problems. The present study deals with the issue of the multi-layered neural networks. It describes selected variants of training algorithms, mainly the standard back-propagation training algorithm and the scaled conjugate gradients algorithm, which ranks among the extremely fast second-order algorithms. One of the parts of the present study is also an application for the visualisation of the structure of multi-layered neural networks whose solution is designed with respect to its potential utilization in the education of artificial intelligence. The first part of the study introduces the subject matter and formally describes both algorithms, followed by a short description of other variants of the algorithms and their analysis. The next part discusses the selection of the appropriate programming language for the implementation of the application, specifies the goals and describes the implementation works. The conclusion summarizes the test results of the speed and implementation comparison with the selected noncommercial-based software ENCOG.
|
|
|
A tool for testing of algorithms for learning languages
Krejčová, Martina ; Hoffmann, Petr (advisor) ; Kukačka, Marek (referee)
Goal of this work was to develop a tool for testing algorithms of grammatical inference of context-free languages. Grammatical inference is a process of learning of grammars and languages from data. Learning could mean finding a suitable model that describes data. Due to this model we could for instance compress this data, create new data or find out which data are consistent with this model. The model in our tool is context-free grammar. We describe how to generate the context-free grammar and data from which the algorithm can try to learn the language of this grammar. The problem of context-free grammar complexity was solved in order to evaluate success achieved by an algorithm in different learning tasks. Also some alternatives of evaluating success of an algorithm are described. This tool is also useful to create data for publishing results of the algorithm.
|
|
|
Artificial neural networks for pattern recognition
Kukačka, Marek ; Mrázová, Iveta (advisor) ; Božovský, Petr (referee)
This work describes the advantages and disadvantages of using neural networks for pattern recognition. Several neural network models are described and their use for pattern recognition is demonstrated. Standard multi-layered perceptron model is compared to a more sophisticated convolutional network model. A new network model is introduced, which is inspired by the convolutional networks and aimed at rectifying some of their shortcomings. The work describes results of tests performed with the described network model on the problem of recognizing hand-written digits.
|
|
|
Artificial neural networks for pattern recognition
Kukačka, Marek
Title: Artificial neural networks for pattern recognition Author: Marek Kukačka Department: Katedra teoretické informatiky a matematické logiky Supervisor: Doc. RNDr. Iveta Mrázová, CSc., Katedra softwarového inženýrství Supervisor's email address: Iveta.Mrazova@mff.cuni.cz Abstract: This work describes the advantages and disadvantages of using neural networks for pattern recognition. Several neural network models are described and their use for pattern recognition is demonstrated. Standard multi-layered perceptron model is compared to a more sophisticated convolutional network model. A new network model is introduced, which is inspired by the convolutional networks and aimed at rectifying some of their shortcomings. The work describes results of tests performed with the described network model on the problem of recognizing hand-written digits. Keywords: neural networks pattern recognition
|
|
|
Multi-layered neural networks and visualization of their structure
Drobný, Michal ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
The model of multi-layered neural networks of the back-propagation type is well-known for their universal approximation capability and even the standard back-propagation training algorithm used for their adjustment often provides results applicable to real-world problems. The present study deals with the issue of the multi-layered neural networks. It describes selected variants of training algorithms, mainly the standard back-propagation training algorithm and the scaled conjugate gradients algorithm, which ranks among the extremely fast second-order algorithms. One of the parts of the present study is also an application for the visualisation of the structure of multi-layered neural networks whose solution is designed with respect to its potential utilization in the education of artificial intelligence. The first part of the study introduces the subject matter and formally describes both algorithms, followed by a short description of other variants of the algorithms and their analysis. The next part discusses the selection of the appropriate programming language for the implementation of the application, specifies the goals and describes the implementation works. The conclusion summarizes the test results of the speed and implementation comparison with the selected noncommercial-based software ENCOG.
|
|
|
RBF-networks with a dynamic architecture
Jakubík, Miroslav ; Mrázová, Iveta (advisor) ; Kukačka, Marek (referee)
In this master thesis I recapitulated several methods for data clustering. Two well known clustering algorithms, concretely K-means algorithm and Fuzzy C-means (FCM) algorithm, were described in the submitted work. I presented several methods, which could help estimate the optimal number of clusters. Further, I described Kohonen maps and two models of Kohonen's maps with dynamically changing structure, namely Kohonen map with growing grid and the model of growing neural gas. At last I described quite new model of radial basis function neural networks. I presented several learning algorithms for this model of neural networks, RAN, RANKEF, MRAN, EMRAN and GAP. In the end of this work I made some clustering experiments with real data. This data describes the international trade among states of the whole world.
|
guest ::
RSS feed.