|
|
Odhadování v úlohách s pravděpodobnostními omezeními
Houda, Michal
Many engineering and economic applications make use of the stochastic programming theory. Major part of models require a complete knowledge of distribution of random parameters, but this assumption is rarely accomplished. We then need to study behaviour of optimal solution when the distribution changes slightly. In our contribution we consider the chance constrained problem;we recapitulated some known theoretical results about stability and estimation of the problem.
|
|
|
Nová kriteria pro stochastiku DEA
Chovanec, Petr
By its nature, Data Envelopment Analysis (DEA) leaves no room for uncertainty in data such as measurement errors. To improve this fact, we consider a-stochastic efficiency concept, and we relate this problem to the stochastic programming problem. Probability inequalities are employed for introducing ew criteria, and two special cases for normal and for general distribution are discussed. The strengths of new criteria are illustrated with a numerical example.
|
|
|
Optimalita prumerne variance v markovskych rozhodovacich procesech
Sladký, Karel ; Sitař, Milan
In this note, we consider discrete-time Markov decision processes with finite state space. Recalling explicit formulas for the growth rate of expected value and variance of the cumulative (random) reward, algorithmic procedures for finding optimal policies with respect to various mean variance optimality criteria are discussed. Computational experience with large scale numerical examples is reported.
|
|
|
Predpovídání ve dvojité aukci se spojitým časem
Šmíd, Martin
Recently, the continuous double auction, i.e. the trading mechanism used in the majority of the financial markets, is the subject of an extensive study. In the present paper, a model of the continuous double auction with the completely random flow of the limit orders is studied. The main result of the paper is an approximate formula for the distribution of the market price and the traded volume at the time s given the information available at t < s.
|
|
| |
|
|
Stabilita úloh stochastického programování s lineární kompenzací
Kaňková, Vlasta
Stochastic programming problems with recourse is a composition of inner and outer optimization problems. A solution of the outer problem depends on the "underlying" probability measure, a solution of inner problem depends on the solution of the outer problem and on the random element realization. Consequently (in the case of the optimal solution of the outer problem) the optimal value and the solution of the inner problem depend also on the probability measure. The aim is to investigate this dependence.
|
|
|
Bayesovská analýza časových řad s kovariátami
Volf, Petr
Bayes methods (supported by MCMC computations) allows to deal with enhanced statistical models. It concerns also time series analysis, where the autoregressive character can be incorporated already to Bayes prior model and one can consider simultaneously a similar time development of other parameters. In present contribution the methodology is used to the analysis of time series of aggregated unemployment data, model contains regression on age, gender, region, and time-dependent variance.
|
guest ::
