Persons

For applications of mathematical statistics in problems of identification of the multidimensional sequences the technique of delimitation of arguments, within which the probability of the admission of a mistake by this or that method is less, than at the competing algorithm, has been created. As subjects to comparison algorithms of formation linear the svertochnykh of codes and hardware observation r - digit vectors in stationary and ergodic process have been chosen. As a result of the executed researches the ratios allowing usage for concrete realization of process the most effective method of identification in terms of the selective probability of a mistake have been received. Scope of applying the of results is the data storage in the computer or information transfer in a communication channel.

• Counter: 947 | Comments : 0

For real applications at the stage of preparing the information for piping an algorithm, having small probability of mistake admission for concrete realization of the formed message, has been chosen. In the paper the method of synthesis for assessment of the probability of registration of the binary (1)-transitions in solving the problems of coding and identification of messages has been considered. The ratio for the probability of a mistake admission, corresponding to the specified method at supervision a priori of uncertain asymptotic selection has been obtained. The fashion of distribution has been defined and the comparative analysis of the probabilities of a mistake admission, characterizing the offered method and the known algorithms of convolution at the given point has been defined. The proposed method of supervision in the probability processes is a more precise algorithm of identification, than the signature analysis.

• Counter: 1110 | Comments : 0