|
|
Analysis of Outlier Detection Methods
Labaš, Dominik ; Bartík, Vladimír (referee) ; Burgetová, Ivana (advisor)
The topic of this thesis is analysis of methods for detection of outliers. Firstly, a description of outliers and various methods for their detection is provided. Then a description of selected data sets for testing of methods for detection of outliers is given. Next, an application design for the analysis of the described methods is presented. Then, technologies are presented, which provide models for described methods of detection of outliers. The implementation is then described in more detail. Subsequently, the results of experiments are presented, which represent the main part of this thesis. The results are evaluated and the individual models are compared with each other. Lastly, a method for accelerating outlier detection is demonstrated.
|
|
|
Data Mining Module of a Data Mining System on NetBeans Platform
Výtvar, Jaromír ; Křivka, Zbyněk (referee) ; Zendulka, Jaroslav (advisor)
The aim of this work is to get basic overview about the process of obtaining knowledge from databases - datamining and to analyze the datamining system developed at FIT BUT on the NetBeans platform in order to create a new mining module. We decided to implement a module for mining outliers and to extend existing regression module with multiple linear regression using generalized linear models. New methods using existing methods of Oracle Data Mining.
|
|
|
Statistical analysis of big industrial data
Zamazal, Petr ; Popela, Pavel (referee) ; Šomplák, Radovan (advisor)
This thesis deals with processing of real data regarding waste collection. It describes select parts of the fields of statistical tests, identification of outliers, correlation analysis and linear regression. This theoretical basis is applied through the programming language Python to process the data into a form suitable for creating linear models. Final models explain between 70 \% and 85 \% variability. Finally, the information obtained through this analysis is used to specify recommendations for the waste management company.
|
|
|
The Introduction and Application of General Regression Model
Hrabec, Pavel ; Štarha, Pavel (referee) ; Bednář, Josef (advisor)
This thesis sumarizes in detail general linear regression model, including testing statistics for coefficients, submodels, predictions and mostly tests of outliers and large leverage points. It describes how to include categorial variables into regression model. This model was applied to describe saturation of photographs of bread, where input variables were, type of flour, type of addition and concntration of flour. After identification of outliers it was possible to create mathematical model with high coefficient of determination, which will be usefull for experts in food industry for preliminar identification of possible composition of bread.
|
|
|
Algorithms for anomaly detection in data from clinical trials and health registries
Bondarenko, Maxim ; Blaha, Milan (referee) ; Schwarz, Daniel (advisor)
This master's thesis deals with the problems of anomalies detection in data from clinical trials and medical registries. The purpose of this work is to perform literary research about quality of data in clinical trials and to design a personal algorithm for detection of anomalous records based on machine learning methods in real clinical data from current or completed clinical trials or medical registries. In the practical part is described the implemented algorithm of detection, consists of several parts: import of data from information system, preprocessing and transformation of imported data records with variables of different data types into numerical vectors, using well known statistical methods for detection outliers and evaluation of the quality and accuracy of the algorithm. The result of creating the algorithm is vector of parameters containing anomalies, which has to make the work of data manager easier. This algorithm is designed for extension the palette of information system functions (CLADE-IS) on automatic monitoring the quality of data by detecting anomalous records.
|
|
|
Algorithms for anomaly detection in data from clinical trials and health registries
Bondarenko, Maxim ; Blaha, Milan (referee) ; Schwarz, Daniel (advisor)
This master's thesis deals with the problems of anomalies detection in data from clinical trials and medical registries. The purpose of this work is to perform literary research about quality of data in clinical trials and to design a personal algorithm for detection of anomalous records based on machine learning methods in real clinical data from current or completed clinical trials or medical registries. In the practical part is described the implemented algorithm of detection, consists of several parts: import of data from information system, preprocessing and transformation of imported data records with variables of different data types into numerical vectors, using well known statistical methods for detection outliers and evaluation of the quality and accuracy of the algorithm. The result of creating the algorithm is vector of parameters containing anomalies, which has to make the work of data manager easier. This algorithm is designed for extension the palette of information system functions (CLADE-IS) on automatic monitoring the quality of data by detecting anomalous records.
|
|
|
Detection of Unusual Events in Temporal Data
Černík, Tomáš ; Bartík, Vladimír (referee) ; Zendulka, Jaroslav (advisor)
Bachelor thesis deals with detection of unusual events (anomalies) in available temporal data. Theoretical part describes existing techniques and algorithms used to detect outliers. There are also introduced meteorological data that are after that used for experimental verification of implemented detection algorithms. Second part, practical one, describes design and implementation of application and algorithms. Algorithms are also tested in search for point, contextual and collective anomalies.
|
|
|
Application Of Implicitly Weighted Regression Quantiles: Analysis Of The 2018 Czech Presidential Election
Kalina, Jan ; Vidnerová, Petra
Regression quantiles can be characterized as popular tools for a complex modeling of a continuous response variable conditioning on one or more given independent variables. Because they are however vulnerable to leverage points in the regression model, an alternative approach denoted as implicitly weighted regression quantiles have been proposed. The aim of current work is to apply them to the results of the second round of the 2018 presidential election in the Czech Republic. The election results are modeled as a response of 4 demographic or economic predictors over the 77 Czech counties. The analysis represents the first application of the implicitly weighted regression quantiles to data with more than one regressor. The results reveal the implicitly weighted regression quantiles to be indeed more robust with respect to leverage points compared to standard regression quantiles. If however the model does not contain leverage points, both versions of the regression quantiles yield very similar results. Thus, the election dataset serves here as an illustration of the usefulness of the implicitly weighted regression quantiles.
|
|
|
The 2020 Election In The United States: Beta Regression Versus Regression Quantiles
Kalina, Jan
The results of the presidential election in the United States in 2020 desire a detailed statistical analysis by advanced statistical tools, as they were much different from the majority of available prognoses as well as from the presented opinion polls. We perform regression modeling for explaining the election results by means of three demographic predictors for individual 50 states: weekly attendance at religious services, percentage of Afroamerican population, and population density. We compare the performance of beta regression with linear regression, while beta regression performs only slightly better in terms of predicting the response. Because the United States population is very heterogeneous and the regression models are heteroscedastic, we focus on regression quantiles in the linear regression model. Particularly, we develop an original quintile regression map, such graphical visualization allows to perform an interesting interpretation of the effect of the demographic predictors on the election outcome on the level of individual states.
|
|
|
Image Deblurring in Demanding Conditions
Kotera, Jan ; Šroubek, Filip (advisor) ; Portilla, Javier (referee) ; Jiřík, Radovan (referee)
Title: Image Deblurring in Demanding Conditions Author: Jan Kotera Department: Institute of Information Theory and Automation, Czech Academy of Sciences Supervisor: Doc. Ing. Filip Šroubek, Ph.D., DSc., Institute of Information Theory and Automation, Czech Academy of Sciences Abstract: Image deblurring is a computer vision task consisting of removing blur from image, the objective is to recover the sharp image corresponding to the blurred input. If the nature and shape of the blur is unknown and must be estimated from the input image, image deblurring is called blind and naturally presents a more difficult problem. This thesis focuses on two primary topics related to blind image deblurring. In the first part we work with the standard image deblurring based on the common convolution blur model and present a method of increasing robustness of the deblur- ring to phenomena violating the linear acquisition model, such as for example inten- sity clipping caused by sensor saturation in overexposed pixels. If not properly taken care of, these effects significantly decrease accuracy of the blur estimation and visual quality of the restored image. Rather than tailoring the deblurring method explicitly for each particular type of acquisition model violation we present a general approach based on flexible automatic...
|
guest ::
