|
|
Design of exercises for data mining - Classification and prediction
Martiník, Jan ; Malý, Jan (referee) ; Burget, Radim (advisor)
My master's thesis on the topic of "Design of exercises for data mining - Classification and prediction" deals with the most frequently used methods classification and prediction. There are association rules, Bayesian classification, genetic algorithms, the nearest method neighbor, neural network and decision trees on the classification. There are linear and non-linear prediction on the prediction. This work also contains a summary of detail the issue of decision trees and a detailed algorithm for creating the decision tree, including development of individual diagrams. The proposed algorithm for creating the decision tree is tested through two tests of data dowloaded from Internet. The results are mutually compared and described differences between the two implementations. The work is written in a way that would provide the reader with a notion of the individual methods and techniques for data mining, their advantages, disadvantages and some of the issues that directly relate to this topic.
|
|
|
ANALYSIS AND DEFINITION OF DECISION PROBLEMS OF EXPERT IN REAL ESTATE VALUATION
Krejza, Zdeněk ; Bradáč, Albert (referee) ; Abraham, Karel (referee) ; Tichá, Alena (advisor)
The thesis deals with the decision-making of the expert in real estate valuation. Due to the complexity of the process and the difficulties of valuation it can be assumed that the decision will be an arduous process. It is obvious that the choice of an expert is crucial to the result of the valuation process. This topic is currently relatively little explored, and therefore the work will deal with the analysis and formulation of decision problems expert in real estate valuation. The thesis analyses the current status of forensic engineering and decision-making regarding to real estate valuation. The general decision-making process, divided into seven steps, is adapted to the requirements of expert decision-making in real estate valuation. As in the managerial decision-making process, property valuation is also divided into three levels. These three levels considered the described fundamental decision problems that lead to the formulation of the expert decision-making principles in real estate valuation. For better understanding the extensiveness of the decision-making process in the valuation of real estate the author created a decision tree respectively schemes whose functionality has been verified at the end of the thesis, exemplified with the help of a specific case study of the determined price in real estate valuation.
|
|
|
Mining Modules of Data Mining System on NetBeans Platform
Henkl, Tomáš ; Lukáš, Roman (referee) ; Zendulka, Jaroslav (advisor)
The master's thesis deals with the knowledge discover in databases and with the extending of the data mining systems in the Oracle environment developed at the VUT FIT. The system kernel conception incorporates an interface that enables the adding of data mining modules. The objective of the thesis is to learn this interface and implement and embed the data mining module for decision-tree classification into the application. In addition, the thesis compares the application with similar commercial product SAS Enterprise Miner
|
|
| |
|
|
Business Intelligence - Use of Data Mining in Business Processes
Skalický, Tomáš ; Veselý, Martin (referee) ; Kříž, Jiří (advisor)
The aim of this bachelor thesis is to get acquainted with the concept of Business Intelligence as well as with the concept of data mining and its use in the company sphere. In the introductory theoretical part I will introduce the tools of Business Intelligence and data mining methods. In the following practical part I will use the methods for analysis of provided company data. The analysis obtained can be used as a support for company decision making.
|
|
| |
|
|
Automatic image screenshots retrieval from video data using JAVA platform
Kulhavý, Miloslav ; Říha, Kamil (referee) ; Burget, Radim (advisor)
This thesis deals with automatic detection of transition scenes videos on the JAVA platform. It was created experiment, which recognizes the scene transitions in video samples and evaluates the recognition accuracy. For the realization of the experiment was created 512 video samples (256 with and 256 without scenes transitions), each of the seven screenshots. These samples were analyzed and by decision tree classified into one of two classes, depending on where they contain a transition of scenes or not. For it was used RapidMiner tool, and its extensions VIMI and IMMI. The purpose of this thesis is train the automatic detection of scenes transition and find the optimal settings of a decision tree for the highest classification accuracy. Highest accuracy was 75.2 %.
|
|
|
Data Mining for Suggesting Further Actions
Veselovský, Martin ; Burget, Radek (referee) ; Bartík, Vladimír (advisor)
Knowledge discovery from databases is a complex issue involving integration, data preparation, data mining using machine learning methods and visualization of results. The thesis deals with the whole process of knowledge discovery, especially with the issue of data warehousing, where it offers the design and implementation of a specific data warehouse for the company ROI Hunter, a.s. In the field of data mining, the work focuses on the classification and forecasting of the advertising data available from the prepared data warehouse and, in particular, on the decision tree classification. When predicting the development of new ads, emphasis is put on the rationale for the prediction as well as the proposal to adjust the ad settings so that the prediction ends positively and, with a certain likelihood, the ads actually get better results.
|
|
|
Decision making based on partially known decision trees
Poláček, Tomáš ; Dostál, Petr (referee) ; Koutský, Jaroslav (referee) ; Váchal, Jan (referee) ; Dohnal, Mirko (advisor)
There is a wide range of different algorithms for insolvency prediction. The complex concept of insolvency proceedings from the point of view of both parties (debtor versus creditor) and from the point of view of the macroeconomics in this dissertation is new. It is often very difficult to generate forecasts using numerical quantifiers and traditional statistical methods. The reason is the lack of input data. Therefore, the work uses trend analysis tools based on the least information intensive quantifiers, ie trends, increasing, constant, and decreasing. A trend model solution is a set of scenarios where a set of variables is quantified by these trends. All possible transitions between the scenarios are generated and plotted in transition graphs. The oriented transition graph has as a node a set of scenarios, and as a branch the transitions between the scenarios. The given path through the transition graph describes any possible future and past behavior of the insolvency system being investigated. The Transition graph is a complete list of trend-based forecasts. The heuristics for determination of the payoff values from the insolvency proceedings applicable to the decision tree tools and the generated transition graphs from trend analyzes are also presented and used in the thesis. A nine-dimensional model serves as a case study. Vague variables are used in models that may have a major impact on the entire insolvency process, eg greed level and political situation.
|
|
|
Identification of Application Protocols
Wrona, Jan ; Bartoš, Václav (referee) ; Kořenek, Jan (advisor)
This thesis is focused on identification of application protocols with emphasizing the speed of their recognition and following possibility of hardware implementation. Nowadays tools are not suitable for fast identification of application protocols in current network monitoring devices, because the decision is not provided for the first packets of network flow. Therefore this thesis propose new model for fast and reliable identification of application protocols. The model was implemented and tested on HTTP, SIP, SMTP and DNS protocols and results were compared to regular expressions and nDPI and libprotoident libraries. For all these protocols, the proposed model has comparable accuracy to other methods, but also provides fast result based on the first packets of the flow.
|
guest ::
