National Repository of Grey Literature 4 records found  Search took 0.01 seconds. 
The Estimation of Probability of Default Using Logistic Regression
Jiřičko, Pavel ; Dlouhá, Zuzana (advisor) ; Formánek, Tomáš (referee)
The aim of the thesis was to build a probability prediction model for client loan repayment. First, the author selected suitable explanatory variables and explained these in detail in the thesis. The author examined various quantities both for qualitative variables (e.g. failure rates and scores) and quantitative variables (e.g. means and standard deviations) describing individual categories. The predictions were done by the use of logistic regression. The estimated coefficients of individual explanatory variables were examined in terms of their statistical significance. The author also described the impact of individual explanatory variables on the probability of a clients default. The probabilities were calculated for two clients with specific values of explanatory variables. The author managed to build a model that can be used to predict probability of client loan repayment defaults.
Management of small business: domestic and foreign experience
Ponomareva, Alla ; Svobodová, Ivana (advisor) ; Chebotar, Yury (referee)
In the first chapter (introduction) revealed the relevance, objectives, purpose, object, object, information database and research methods. The second chapter covers theoretical aspects of small business management case studies, including the history of small business development in Russia, General characteristics and basic principles of small business management, State small business development in Russia and abroad. The third chapter examines the current state and problems of management of development of small entrepreneurship in a major metropolis. Analysis and evaluation of the Office of small business development in a major metropolis. Developed guidelines for improving the management of development of small business in today's environment, including the use of foreign experience for small business development in Russia and the development strategy of the State support and development of small and medium-sized businesses in a major metropolis. This strategy may include: an integrated approach to small business development, creating sustainable infrastructure to support small businesses, increasing entrepreneurial goramotnosti, involvement of young people in entrepreneurship and a number of other areas of the State support of small and medium-sized businesses in a major metropolis.
Performance Evaluation of Graphics Accelerator
Vanek, Juraj ; Drábek, Vladimír (referee) ; Herout, Adam (advisor)
My work is about performance evaluation of graphics accelerator by 3D scenes rendering. It focuses on how to test the performance of the accelerator from every point of view during the performance critical operations in 3D graphics. It deals with implementation problems of individual measures, their influence on total results and evaluation of the measurements. The final output of the application is a score, by which is possible to judge what performance is giving the accelerator in comparison with others. For this comparison is available an on-line results database.
Performance Evaluation of Graphics Accelerator
Vanek, Juraj ; Jošth, Radovan (referee) ; Herout, Adam (advisor)
This paper deals with possibilities and functions of modern graphic accelerators and with measuring performance under OpenGL interface. Widespread algorithms to render scene in real-time are used. It focuses on how to test every part of accelerator's graphic pipeline as well as measure performance in rendering of advanced effects and theoretical speed at general purpose calculations through graphic processor. This testing is realized by implementing multiple test series and their further evaluation. Final application enables setting of test parameters and outputs a score, by which is possible to judge accelerator's performance in comparison among themselves.

Interested in being notified about new results for this query?
Subscribe to the RSS feed.