Příspěvky z konference

Nejnovější přírůstky:
2018-06-19
19:26

Úplný záznam
2018-06-19
19:26

Úplný záznam
2018-06-19
19:26
Gradient Descent Parameter Learning of Bayesian Networks under Monotonicity Restrictions
Plajner, Martin ; Vomlel, Jiří
Learning parameters of a probabilistic model is a necessary step in most machine learning modeling tasks. When the model is complex and data volume is small the learning process may fail to provide good results. In this paper we present a method to improve learning results for small data sets by using additional information about the modelled system. This additional information is represented by monotonicity conditions which are restrictions on parameters of the model. Monotonicity simplifies the learning process and also these conditions are often required by the user of the system to hold. \n\nIn this paper we present a generalization of the previously used algorithm for parameter learning of Bayesian Networks under monotonicity conditions. This generalization allows both parents and children in the network to have multiple states. The algorithm is described in detail as well as monotonicity conditions are.\n\nThe presented algorithm is tested on two different data sets. Models are trained on differently sized data subsamples with the proposed method and the general EM algorithm. Learned models are then compared by their ability to fit data. We present empirical results showing the benefit of monotonicity conditions. The difference is especially significant when working with small data samples. The proposed method outperforms the EM algorithm for small sets and provides comparable results for larger sets.

Úplný záznam
2018-06-19
19:26
Employing Bayesian Networks for Subjective Well-being Prediction
Švorc, Jan ; Vomlel, Jiří
This contribution aims at using Bayesian networks for modelling the relations between the individual subjective well-being (SWB) and the individual material situation. The material situation is approximated by subjective measures (perceived economic strain, subjective evaluation of the income relative to most people in the country and to own past) and objective measures (household’s income, material deprivation, financial problems and housing defects). The suggested Bayesian network represents the relations among SWB and the variables approximating the material situation. The structure is established based on the expertise gained from literature, whereas the parameters are learnt based on empirical data from 3rd edition of European Quality of Life Study for the Czech Republic, Hungary, Poland and Slovakia conducted in 2011. Prediction accuracy of SWB is tested and compared with two benchmark models whose structures are learnt using Gobnilp software and a greedy algorithm built in Hugin software. SWB prediction accuracy of the expert model is 66,83%, which is significantly different from no information rate of 55,16%. It is slightly lower than the two machine learnt benchmark models.

Úplný záznam
2018-06-19
19:26
Dynamic Bayesian Networks for the Classification of Sleep Stages
Vomlel, Jiří ; Kratochvíl, Václav
Human sleep is traditionally classified into five (or six) stages. The manual classification is time consuming since it requires knowledge of an extensive set of rules from manuals and experienced experts. Therefore automatic classification methods appear useful for this task. In this paper we extend the approach based on Hidden Markov Models by relating certain features not only to the current time slice but also to the previous one. Dynamic Bayesian Networks that results from this generalization are thus capable of modeling features related to state transitions. Experiments on real data revealed that in this way we are able to increase the prediction accuracy.

Úplný záznam
2018-06-19
19:26
Tourist interest in illicit zone of ice caves
Nováková, Eva ; Kuda, František ; Kubalíková, Lucie
Ledové sluje (Ice Caves) in the Podyjí National Park represents one of the most spectacular sites within the area. It consists of the large boulder field and several pseudokarst caverns on the north-western slope of the ridge that are very important from the ecological and geomorphological point of view. The access to the site is restricted for the visitors of National Park as there exists a risk of damage and disturbance of these unique phenomena; tourist can use the marked paths leading around the site, they can reach the top part of the ridge.\nCurrently, there are only several persons who have legal access to the site (employees of the National Park Administration and other researches with the permission issued by NP Administration). However, the installed sensor that counts the passages proved that the site is visited more frequently than it should be. The number of people who visit this site (situated within the first zone of National Park where there is no marked path and so the access is forbidden by decree) is quite alarming. Based on these findings, some proposals for the solution of this unfavourable situation are proposed and other possibilities how to avoid this undesirable phenomenon are discussed.\n
Plný text: Stáhnout plný textPDF

Úplný záznam
2018-06-19
19:26
Použití modelu efektivní trhliny k analýze odezvy válcových těles se šípovým zářezem
Halfar, P. ; Frantík, P. ; Šimonová, H. ; Daněk, P. ; Keršner, Z. ; Vavro, Leona ; Vavro, Martin
Příspěvek uvádí aplikaci modelu efektivní trhliny k analýze odezvy válcových těles se\nšípovým zářezem namáhaných v tříbodovém ohybu. Jednalo se o testy tří zkušebních\ntěles z pískovce z lokality Javorka a tří betonových těles z nosné konstrukce nádražní\nbudovy v Ostravě-Vítkovicích. K výpočtům byl použit akademický software\nchevroncylinder na bázi metody konečných prvků.

Úplný záznam
2018-06-19
19:26
Eternal be the Memory of our Lady: the Synodikon of Orthodoxy and the Piety of the Late Byzantine Empresses
Melichar, Petra
The study focuses on the fact that only nine out of fifteen late Byzantine empresses were confirmed by the Orthodox Church as Orthodox and their names were inscribed in the Synodikon of Orthodoxy (which among others contains a list of names of Byzantine rulers who died in Orthodoxy). The author inquires into the cases of empresses whose names were not included and suggests the reasons for their omission. The results of the study reveal the criteria of evaluation set by the representatives of the Orthodox Church for the official piety of empresses.

Úplný záznam
2018-06-19
19:26
Method Development for Speciation Analysis of Volatile Mercury Forms in Ambient Air and flue Gases from Incineration Plants.
Švehla, Jaroslav ; Kratzer, J. ; Svoboda, M. ; Svoboda, Karel ; Šyc, Michal ; Ružovič, Tomáš
Potential of AAS for on-line speciation analysis of gaseous Hg0 (non-heated atomizer) and HgCl2 (heated atomizer) in flue gases is discussed.
Plný text: Stáhnout plný textPDF

Úplný záznam
2018-06-19
19:26
Alternative Methods of Permeation Barriers Preparation
Veverka, Jakub ; Matějíček, Jiří
Permeation barriers represent one of the crucial fields in materials development for thermonuclear fusion. Primary objective of the barriers is to suppress the permeation of hydrogen isotopes (mainly tritium) from future thermonuclear fusion facilities. Secondary objective is to reduce their retention in structural materials. Expected reactor conditions put high demands on the material, as well as on the final barrier quality. Key properties are tritium permeation reduction, absence of defects (especially cracks), high-temperature stability and corrossion resistance, and compatibility with structural materials (mostly ferritic-martensitic steels). Thin nitride layers, identified as promising permeation barriers, were prepared by diffusion-based nitridation and physical vapour deposition (PVD), and characterized.

Úplný záznam