National Repository of Grey Literature 94 records found  beginprevious85 - 94  jump to record: Search took 0.00 seconds. 
Artificial neural networks for clustering and rule extraction
Iša, Jiří ; Jiroutek, Pavel (referee) ; Mrázová, Iveta (advisor)
Rule extraction with neural networks has been a common research topic over the last decades. This master thesis proposes a novel growing fuzzy inference neural network, based on the principle of growing neural structures [5]. This allows the network to adjust iteratively its number of hidden neurons. For the purpose of this network an existing clustering algorithm is enhanced to improve the sensitivity to the requested output. A novel fast weights adaptation, inspired by the fuzzy set theory, is also suggested. The characteristics of the proposed model and a new method of the selection of significant input features support the induction of a relatively small amount of simple fuzzy rules. The introduced techniques have been experimentally tested on real-world data describing the relationship between various types of housing in the Boston area and its price. The data was obtained from the "Boston housing" dataset.
Decision Trees and Knowledge Extraction
Vitinger, Jiří ; Mrázová, Iveta (advisor) ; Jiroutek, Pavel (referee)
The goal of data mining is to extract knowledge, dependencies and rules from data sets. Many complex methods were developed to solve it. This thesis presents some of the most important methods, which include the decision trees with algorithms ID3, C4.5 and CART, neural networks like multilayer neural networks with the backpropagation algorithm, RBF networks, Kohonens maps and some modifications of LVQ method. There are also described some clustering methods like hierarchical clustering, QT clustering, kmeans method and its fuzzy modification. The work also includes data pre-processing techniques, which are very important in order to obtain better results of data mining process. Experimental part of the work compares the presented methods by means of the results of many tests on real-world data sets. The results can be used as a guide to choose an appropriate method and its parameters for some given data set. In this work there is presented author's implementation of the decision trees C4.5 and CART in C#. In the application it is possible to watch details of algorithms work. The application provides an API enabling an implementation of new algorithms.
Recognition of structured noises by neural network synchronization
Krchák, Jakub ; Mrázová, Iveta (referee) ; Maršálek, Petr (advisor)
This work studies the phenomenon of sound recognition through spiking neuron network synchronization. The input layer res on specifi c features in the input sound, which resemble syllable. The neurons in the middle layer are interconnected in such a way that they prolong their ring rates if the ring frequence is similar. This causes the ring of the output neuron of the corresponding pattern.
Experimental GUHA procedures
Kuchař, Tomáš ; Mrázová, Iveta (referee) ; Rauch, Jan (advisor)
The goal of this work is a new implementation of six GUHA procedures known from LISpMiner system (4ftMiner, SD4ft-Miner, CFMiner, SDCFMiner, KLMiner, SDKLMiner) into Ferda Data Miner system environment with respect to their futher research and development. GUHA procedure automatically generates patterns from user defined set of relevant patterns and tests if it is true in the analysed data. The output of the procedure consists of all prime patterns. The pattern is prime if it is true in the analysed data and if it does not immediately follow from the other more simple output patterns. Typical effective implementation of a GUHA procedure uses suitable database representation by bit strings. Tools were created for solving above-mentioned GUHA procedures. During works were extended options of entering the relevant questions set.
Artificial neural networks and their application in the assessment of insurance risks
Macek, Karel ; Jiroutek, Pavel (referee) ; Mrázová, Iveta (advisor)
The present work studies applicability of artificial neural networks in the assessment of insurance risk. Structure and function of multilayer perceptrons, Kohonen maps, Fuzzy ART networks, and Fuzzy ARTMAP networks are described. Concept of insurance risk is defined and the ratemaking by generalized linear models is introduced. Neural networks' methods for reduction of input space's dimension, knowledge extraction, and visualization are summed up. Data describing traffic accidents are acquainted and results achieved on them are presented. The work successfully demonstrates theoretically and experimentally that multilayer perceptrons approximate better than generalized linear models. Modification of multilayer perceptron estimated distribution function of total claim. Analysis performed by Kohonen map and by subsequent visualization detected two significant clusters. Analysis by Fuzzy ART network is presented as a part of new algorithm for reduction of input space's dimension. New algorithm inspired by Fuzzy ARTMAP network discovered a group of accidents where the claim is above average. This group is delimited by interpretable rules. Attached CD contains scripts for Matlab and MySQL that was used for mentioned analyses.
Feed-forward neural networks and their application in data mining
Civín, Lukáš ; Štanclová, Jana (referee) ; Mrázová, Iveta (advisor)
The goal of data mining is to solve various problems dealing with knowledge extraction from huge amounts of real-world data, the quality of which might be disputable. Neural networks can help with the solution due to their generalization capabilities. While working on data mining projects, we have essentially the following two objectives in real-world applications of feed-forward neural networks. To obtain applicable results, it is crucial to provide the networks with well-prepared data. However, it is equally important to choose the right training strategy for the networks themselves - including network architecture, parameter settings or the training algorithm. One of the most important ideas behind these steps is namely to prevent "over-training". The final network should recall unknown examples as well as possible. There are plenty of techniques with different approaches to the solution. It is possible to modify the data, these comprises modifying the range of the data or its dimension, adding noise to the data, etc. Yet another way is the modification of the neural network by structural learning with forgetting, weight decay or early stopping. These techniques are analyzes both theoretically and experimentally in this thesis. With regard to the results achieved in a number of experimental tests we have...
On-line learning in real-time environments
Pacovský, Ondřej ; Spier, Emmet (advisor) ; Mrázová, Iveta (referee)
In this work, a novel reinforcement learning algorithm, Stimulus Action Reward Network (SARN), is developed. It is targeted for application in real-time domainswhere the inputs are usually continuous and adaptation must proceed on-line, without separate training periods. Another objective is to minimise the amount of problem-specific teacher(human) input needed for successful application of the algorithm. The SARN architecture combines a connectionist network and scalar reinforcement feedback by employing Hebbian principles. By adapting the network weights, connections are established between stimuli and actions that lead to positive feedback. Since the links between the input stimuli and the actions are formed quite rapidly, it is possible to use a large number of stimuli. This leads to the idea of using recurrent random network (Echo State Network) as a pre-processing layer. Prototype implementation is tested in Unreal 2004 game environment. The comparison with Q-learning shows that on the time scale of tens of seconds to minutes, SARN typically achieves better performance. When coupled with an Echo State Network, SARN requires a uniquely low amount of problem-specific information supplied by the teacher. These features make SARN useful for domains such as autonomous robot control and game AI.
Level of Quality Management and Safety Survey in Selected Hospitals in the Czech Republic.
MRÁZOVÁ, Iveta
Quality of care has become an important concept in present nursing. Accreditation is proper evidence and also the most efficient means to achieve high-quality care. Procedure of accreditation in the Czech Republic is most frequently administered at national level by the Joint Commission on Accreditation of the Czech Republic. The main objective of my thesis was to survey the setting of quality management and safety in selected hospitals in the Czech Republic. I used quantitative research and quantitative method {--} interviews and questionnaires in hospitals in České Budějovice, Český Krumlov, Písek, Prachatice, Tábor, Jindřichův Hradec from January to March in 2010. The qualitative part had the form of interviews with staff-nurses from above mentioned hospitals. For this part of work I set seven survey questions that had stemmed from the standards of accreditation of the Joint Commission on Accreditation. These questions referred to implementation of Programme of Quality Improvement and Safety of Provided Services, to the process of auditing, to the quality and safety data collecting, to the process of observing clients/patients´ satisfaction with nursing care, to the setting of report system of unusual and abnormal situations, to the setting of efficient system of complaints solving and to the standardization of health care. On the grounds of these interviews 9 hypotheses were set. These hypotheses were acknowledged in the quantitative survey among head nurses from above mentioned hospitals. 111 questionnaires were handed out, 82 questionnaires returned. The first hypothesis (proved) was: Head nurses play active parts in the Programme of Quality Improvement and Safety of Services Provided. The second hypothesis (not proved) was: Head nurses who perform the function of an auditor, have the certificate of accredited course. The third hypothesis (proved) is: The most frequent indicator of quality of nursing care is observation of patients´ satisfaction with provided care. The fourth hypothesis (proved) is: Head nurses have a form for reporting unusual/abnormal situations available. The fifth hypothesis (proved) is: Patients/clients have questionnaires available to express satisfaction with the care. The sixth hypothesis (proved) is: Corrective measures are taken at wards, on the grounds of patients/clients´ satisfaction with care. The seventh hypothesis (proved): Head nurses participate in preparation of nursing care standards. The eighth hypothesis (not proved) is: Head nurses engage nurses with advanced education in preparation of the standards. The results proved that only those nurses are engaged, who display interest, regardless of their education. The ninth hypothesis (proved) is: Nursing care provided at individual wards has a form of nursing process. Neither nursing process nor nursing documentation is used in one hospital only. The results of the survey will be provided to staff-nurses of the hospitals as a background for continuous quality improvement of provided care.
The contentment with nursing care at children´s department from the view of a person accomplanying the child
MRÁZOVÁ, Iveta
Satisfaction monitoring is one of the indicators of quality of provided care. The quality of nursing care provided in hospitals has been largely discussed recently. For many health care facilities the patients´ opinion is crucial for the nursing care evaluation and represents a feedback for the nursing staff. The objective of my thesis was to examine opinions of persons accompanying child patients in hospital on the nursing care quality. I set three hypotheses: The first hypothesis is {--} persons accompanying child patients in hospital are satisfied with the care, the second states {--} persons accompanying child patients in hospital are satisfied with communication with the nursing staff in a paediatric ward and the third hypothesis suggests that persons accompanying child patients in hospital are not satisfied with the paediatric ward equipment. The research sample was formed by persons accompanying child patients in the Český Krumlov hospital,Inc. Those persons were staying is hospital together with children during their hospitalization. The research was conducted in the paediatric ward of the Český Krumlov hospital,Inc. by the questionnaire method. The research took place in January, February and March. I analysed data obtained from 69 filled in questionnaires. The data analysis shows that persons accompanying children in hospital were satisfied with the care provided {--} the hypotesis I is confirmed. The persons accompanying children in hospital were also satisfied with communication {--} the hypothesis II is confirmed, and they were satisfied with the paediatric ward equipment {--} the hypothesis III isn´t confirmed. The results of my research will be given over to the Český Krumlov hospital, Inc. management, primarily to the Chief Nursing Officer of the Hospital, to the Chief Physician and to the Senior Nurse Manager of the paediatrical ward. The results may serve as a feedback and an expression of respect for the quality of nursing care in the paediatric ward in the hospital Český Krumlov,Inc.

National Repository of Grey Literature : 94 records found   beginprevious85 - 94  jump to record:
See also: similar author names
4 MRÁZOVÁ, Ivana
1 Mrázová, I.
4 Mrázová, Iva
4 Mrázová, Ivana
Interested in being notified about new results for this query?
Subscribe to the RSS feed.