|
|
Adaptive Matchmaking Algorithms for Computational Multi-Agent Systems
Kazík, Ondřej ; Neruda, Roman (advisor) ; Paprzycki, Marcin (referee) ; Diamantini, Claudia (referee)
The multi-agent systems (MAS) has proven their suitability for implementation of complex software systems. In this work, we have analyzed and designed the data mining MAS by means of role-based organizational model. The organiza- tional model and the model of data mining methods have been formalized in the description logic. By matchmaking which is the main subject of our research, we understand the recommendation of computational agents, i.e. agents encap- sulating some computational method, according their capabilities and previous performances. The matchmaking thus consist of two parts: querying the ontol- ogy model and the meta-learning. Three meta-learning scenarios were tested: optimization in the parameter space, multi-objective optimization of data min- ing processes and method recommendation. A set of experiments in these areas have been performed. 1
|
|
|
Social networks and data mining
Zvirinský, Peter ; Mrázová, Iveta (advisor) ; Neruda, Roman (referee)
Recent data mining methods represent modern approaches capable of analyzing large amounts of data and extracting meaningful and potentially useful information from it. In this work, we discuss all the essential steps of the data mining process - including data preparation, storage, cleaning, data analysis as well as visualization of the obtained results. In particular, this work is focused on the data available publicly from the Insolvency Register of the Czech Republic, that comprises all insolvency proceedings commenced after 1. January 2008 in the Czech Republic. With regard to the considered type of data, several data mining methods have been discussed, implemented, tested and evaluated. Among others, the studied techniques include Market Basket Analysis, Bayesian networks and social network analysis. The obtained results reveal several social patterns common in the current Czech society.
|
|
|
Adaptive Matchmaking Algorithms for Computational Multi-Agent Systems
Kazík, Ondřej ; Neruda, Roman (advisor) ; Paprzycki, Marcin (referee) ; Diamantini, Claudia (referee)
The multi-agent systems (MAS) has proven their suitability for implementation of complex software systems. In this work, we have analyzed and designed the data mining MAS by means of role-based organizational model. The organiza- tional model and the model of data mining methods have been formalized in the description logic. By matchmaking which is the main subject of our research, we understand the recommendation of computational agents, i.e. agents encap- sulating some computational method, according their capabilities and previous performances. The matchmaking thus consist of two parts: querying the ontol- ogy model and the meta-learning. Three meta-learning scenarios were tested: optimization in the parameter space, multi-objective optimization of data min- ing processes and method recommendation. A set of experiments in these areas have been performed. 1
|
|
| |
|
|
Empirical comparison of systems for knowledge discovery in databases
Benešová, Kristýna ; Berka, Petr (advisor) ; Šimůnek, Milan (referee)
S rostoucím množstvím shromažďovaných a ukládaných dat roste také potřeba a zájem majitelů těchto dat o využití jejich potenciálu k dalšímu rozhodování. Proto se vyvíjí nové přístupy a způsoby vycházející z informatiky, statistiky a oblasti strojového učení, které se této potřebě snaží vyhovět. Cílem této diplomové práce je uvést proces dobývání znalostí dat z databází na medicínských datech Tinnitus a představit systémy LISp-Miner a Weka, které daný proces podporují. Obsahem teoretické části diplomové práce je shrnutí základních charakteristik a přístupů procesu dobývání znalostí. Praktická část diplomové práce je věnována realizaci celého procesu v jednotlivých krocích. V samotném kroku modelování jsou využity již zmíněné systémy akademické LISp-Miner a Weka. Poslední část praktické části práce patří prezentaci dosažených výsledků a vlastnímu zhodnocení systémů.
|
|
|
Empirical comparison of free software suites for knowledge discovery from data
Kasík, Josef ; Berka, Petr (advisor) ; Rauch, Jan (referee)
Both topic and main objective of the diploma thesis is a comparison of free data mining suites. Subjects of comparison are six particular applications developed under university projects as experimental tools for data mining and mediums for educational purposes. Criteria of the comparison are derived from four general aspects that form the base for further analyses. Each system is evaluated as a tool for handling real-time data mining tasks, a tool supporting various phases of the CRISP-DM methodology, a tool capable of practical employment on certain data and as a common software system. These aspects bring 31 particular criteria for comparison, evaluation of whose was determined by thorough analysis of each system. The results of comparison confirmed the anticipated assumption. As the best tool the Weka data mining suite was evaluated. The main advantages of Weka are high number of machine learning algorithms, numerous data preparation tools and speed of processing.
|
|
|
Data Mining of Macroeconomic Data
Lang, Lukáš ; Berka, Petr (advisor) ; Marek, Luboš (referee)
The theme of my work is the Data Mining (DM) of Macroeconomic Data. The purpose of this work is to use DM methods for analysis of macroeconomic fundamentals of selected countries of the Western Europe and the U.S.A. between years 1961-1989 and to compare the DM methods with statistical methods. For the statistical analysis, I used EViews and MS-Office Excel, for the DM I used LISp-Miner. The structure of the work is as follows: in the theoretical part I define the analysed indicators and their relations with respect to the history of analysed period. Then are specified chosen statistical methods also with the reason for choice. The last chapter of the theoretical part describes the data mining. In the practical part I describe the problems which I solved, data collecting and preparation, use of the statisitical methods and DM methods and results obtained. The enlightenment lies in conclusion. I thank my supervisor, Prof. Ing. Petr Berka, CSc. for DM meditations and important suggestions. I thank my colleagues, Ing. Vojtech Menzl, MSc and Mgr. Jana Závacká for critique of the statistical methods. I thank the living members of my family for patience.
|
guest ::
