Journal:Informatica
Volume 9, Issue 4 (1998), pp. 425–436
Abstract
Principles of the framework called time series forecasting automation are presented. It is required in processing massive temporal data sets and creating completely user-oriented forecasting software where manual data analysis and a user's decision-making is either impractical or undesirable. Its distinct features are local extrapolation models, their active training, criterion of model performance assessment used in adding new examples to the model training set and in deciding on which one of a group of competing models consistent with the common training set performs best. A generalized algorithm for local model tuning on massive data series that can be run without human intervention is presented.
Journal:Informatica
Volume 9, Issue 4 (1998), pp. 415–424
Abstract
Comparative study of the recognition of nonsemantic geometrical figures by the human subjects and ART neural network was carried out. The results of computer simulation experiments with ART neural network showed well correspondence with the psychophysical data on the recognition of different complexity visual patterns: in both cases the patterns of medium complexity were recognized with the highest accuracy. On the contrary, the recognition of the patterns by their informative fragments demonstrated different recognition strategies employed by natural and artificial neural systems. For biological systems, it is necessary the presence of not only distinctive features in visual patterns but the redundant features as well for successive recognition. ART neural network ignores redundant features and recognizes visual patterns with equal accuracy whether the whole pattern or only the informative fragment of any completeness is present.
Journal:Informatica
Volume 9, Issue 4 (1998), pp. 401–414
Abstract
The sample-based rule obtained from Bayes classification rule by replacing unknown parameters by ML estimates from stratified training sample is used for classification of random observations into one of two widely applicable Gamma distributions. The first order asymptotic expansions of the expected risk regret for different parametric structure cases are derived. These are used to evaluate performance of the proposed classification rule and to find the optimal training sample allocation minimizing the asymptotic expected risk regret.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 387–395
Abstract
Laplace equations were used for modeling of electrotonic potential in three-dimensional isotropic double-space RC medium. Solutions of Laplace equations for the case of rectangular current pulse stimulation using spherical electrode we obtain using Laplace transforms in imaginary space. Solutions in original space we got using numerical invert Laplace transform. It showed that the rising front of the transmembrane potential becomes less steep in regard to rising radius of the stimulating electrode and asymptotically reaches single-dimensional cable case (evenly distributed RC-circuit). The steady state value of transmembrane potential decreases with the increasing distance from stimulating electrode. It remains always positive when stimulus current is negative.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 365–386
Abstract
We discuss an age-sex-structured population dynamics deterministic model taking into account random mating of sexes, females' pregnancy and its dispersal in whole space. This model can be derived from the previous one (Skakauskas, 1995) describing migration mechanism by the general linear elliptic operator of second order and includes the male, single (nonfertilized) female and fertilized female subclasses. Using the method of the fundamental solution for the uniformly parabolic second-order differential operator with bounded Hölder continuous coefficients we prove the existence and uniqueness theorem for the classic solution of the Cauchy problem for this model. In the case where dispersal moduli of fertilized females are not depending on age of the mated male we analyze population growth and decay.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 343–364
Abstract
This paper describes a preliminary algorithm performing the mapping of sound to music score. Our procedure is constructed over signal-extracted energy and fundamental frequency traces alone. The algorithm is tested on real songs of average complexity. Although results seem to be promising, their detailed examination reveals some shortages of our approach as well as the set of application specific problems. It appears that musical analysis can not be entirely dissociated from phonetic processing. Further work should be oriented towards integration of knowledge of music as well.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 325–342
Abstract
In the previous papers (Masreliez and Martin, 1977; Novovičova, 1987; Schick and Mitter, 1994) the problem of recursive estimation of linear dynamic systems parameters and of the state of such systems in the presence of outliers in observations have been considered. In this connection various ordinary recursive techniques are worked out, when systems output is corrupted by an additive noise with a time homogeneous contamination of outliers. The aim of the given paper is the development of an approach for robust recursive state estimation of linear dynamic systems in a case of additive noises with time-varying outliers. The recursive technique based on the abovementioned theoretical results is obtained and proved by state estimation of the real chemical process (Box and Jenkins, 1970). The results of numerical simulation by computer (Fig. 1–3) are given.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 315–324
Abstract
We study invertibility of big n × n matrices. There exists a number of algorithms, especially in mathematical statistics and numerical mathematics, requiring to invert step by step large matrices which are closely related to each other. Standard inverting methods require O(n3) arithmetical operations therefore using of these algorithms for big values of n becomes problematic. In this paper we introduce some classes of matrices that can be inverted by O(n2) operations if we use inverse matrices of other closely related matrices. The most important among them are matrices having big common submatrix and modified sample covariance matrices. We apply our theoretical results constructing a fast algorithm for prediction. This algorithm demonstrates the advantage of our inverting methods and can be used, for example, for safety control in the plant.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 297–314
Abstract
More real systems have many components and their simulation requires significant execution times. The practical needs have conducted to distributed simulation rather than sequential method. Asynchronous parallel discrete event simulation (PDES) is studied and its methodology is presented. The paper presents the conservative methodology of PDES and illustrates it with a suggestive particular application: Virtual Assembly Cells.
Journal:Informatica
Volume 9, Issue 3 (1998), pp. 279–296
Abstract
This work is a continuation of our previous papers devoted to exploration of the regular search procedures efficiency in binary search spaces. Here we formulate the problem in a rather general form as a problem of optimization of an unimodal pseudoboolean function given implicitly and obtain analitical estimates of the expected time of a minimum point search for procedures of direct local search. These estimates are polinomial for the case of weakly nonmonotone functions and exponential for the general case of arbitrary unimodal functions. We hope that the proposed result will be usefull first of all for practical applications.