Open AccessFeatures of construction of evaluation algorithms multidimensional random processes
Author(s) -
А. А. Лобатый,
А. Ю. Бумай
Publication year - 2020
Publication title -
sistemnyj analiz i prikladnaâ informatika
Language(s) - English
Resource type - Journals
eISSN - 2414-0481
pISSN - 2309-4923
DOI - 10.21122/2309-4923-2020-1-24-32
Subject(s) - verisimilitude , a priori and a posteriori , dimension (graph theory) , algorithm , process (computing) , multivariate random variable , probabilistic logic , computer science , mathematical optimization , mathematics , stochastic process , maximum a posteriori estimation , random variable , statistics , artificial intelligence , maximum likelihood , philosophy , epistemology , pure mathematics , operating system
The problem of evaluating the information which is present in random signals from various sources-meters is considered. It is assumed that the random process which is evaluated and the meter output according mathematical description of the problem are vector random processes. In this case, the dimension of the vector of the measurement can be larger than the dimension of the vector of the process being evaluated. The analysis of analytical methods and algorithms of the estimation that based on the determination of the main probabilistic characteristics of a random process by both the a priori and the a posteriori methods with various optimality criteria is carried out. Based on the analysis, the problem of complexing of the meters of the random process is considered according the proposed criterion for the maximum of posterior verisimilitude, combining the criterion of maximum verisimilitude and the criterion of maximum the a posteriori probability, general methodology complexing is developed. Proposed example of the complexing shows the efficiency of the proposed method. This approach to the construction of the algorithms of the evaluation for multidimensional random processes allows to increase the accuracy of estimation, since it takes into account additional information and its complex processing.