|
|
Methods for Clustering Data
Pohlídal, Antonín ; Burgetová, Ivana (referee) ; Bartík, Vladimír (advisor)
This bachelor's thesis deals with hierarchical clustering methods with a focus on implementation of agglomerative hierarchical clustering method and its comparison with the DENCLUE method. First of all, various methods are described with emphasis on hierarchical clustering methods. Further, there is an implementation of the selected method, using the Java programming language and MySQL database. The last part contains a comparison with the implementation of DENCLUE method, implemented by Mr. Bc. Radim Kapavík.
|
|
|
Computer Library with Clustering Methods
Riša, Martin ; Homoliak, Ivan (referee) ; Košík, Michal (advisor)
The aim of this work is to create a library with chosen clustering methods, to compare their effectiveness and their properties by testing them on different input data sets. The aim of the testing is to determine efficiency of a method, to determine advantages and disadvantages of a method to cluster general input data or to cluster only data of specific shapes. Stages of development of the library are also documented in the text of this work.
|
|
|
Success Rate Measure Methods in Data Mining
Trunkát, Jan ; Zelený, Jan (referee) ; Bartík, Vladimír (advisor)
The Bachelor thesis is aimed at success rate measure methods in data mining in the area of clustering. It introduces the basic concepts, features of data mining and especially the cluster analysis. This work includes program, which implements methods of measuring success. In conclusion, they are given results of clustering success.
|
|
|
Precipitation Forecast from Weather Radar
Tábi, Matej ; Milička, Martin (referee) ; Burget, Radek (advisor)
The aim of this bachelor thesis is a web application, which solves a short-time precipitation forecast for Czech Republic. The forecast is created for the next one hour and consists of four radar frames with a 15 minutes distance between them. The task is solved in three main steps. In the first step, the actual information from CHMI webpages is obtained. In the second step, the recieved data is processed - precipitation fields are identified and in the last step, movement vectors of these precipitation fields are calculated. The result of the thesis is a fully working web application. The forecast success rate is at the level from 30 to 60%.
|
|
|
Computer Library with Clustering Methods
Riša, Martin ; Homoliak, Ivan (referee) ; Košík, Michal (advisor)
The aim of this work is to create a library with chosen clustering methods, to compare their effectiveness and their properties by testing them on different input data sets. The aim of the testing is to determine efficiency of a method, to determine advantages and disadvantages of a method to cluster general input data or to cluster only data of specific shapes. Stages of development of the library are also documented in the text of this work.
|
|
|
Precipitation Forecast from Weather Radar
Tábi, Matej ; Milička, Martin (referee) ; Burget, Radek (advisor)
The aim of this bachelor thesis is a web application, which solves a short-time precipitation forecast for Czech Republic. The forecast is created for the next one hour and consists of four radar frames with a 15 minutes distance between them. The task is solved in three main steps. In the first step, the actual information from CHMI webpages is obtained. In the second step, the recieved data is processed - precipitation fields are identified and in the last step, movement vectors of these precipitation fields are calculated. The result of the thesis is a fully working web application. The forecast success rate is at the level from 30 to 60%.
|
|
|
Success Rate Measure Methods in Data Mining
Trunkát, Jan ; Zelený, Jan (referee) ; Bartík, Vladimír (advisor)
The Bachelor thesis is aimed at success rate measure methods in data mining in the area of clustering. It introduces the basic concepts, features of data mining and especially the cluster analysis. This work includes program, which implements methods of measuring success. In conclusion, they are given results of clustering success.
|
|
|
Methods for Clustering Data
Pohlídal, Antonín ; Burgetová, Ivana (referee) ; Bartík, Vladimír (advisor)
This bachelor's thesis deals with hierarchical clustering methods with a focus on implementation of agglomerative hierarchical clustering method and its comparison with the DENCLUE method. First of all, various methods are described with emphasis on hierarchical clustering methods. Further, there is an implementation of the selected method, using the Java programming language and MySQL database. The last part contains a comparison with the implementation of DENCLUE method, implemented by Mr. Bc. Radim Kapavík.
|
guest ::
