Оптимизация вычислений при применении генетических алгоритмов в задаче структурно-параметрической идентификации регрессионных моделей
The computationally efficient method of fitness function evaluation (criterion for chromosomes selection) in genetic algorithms (GA) is discussed in this paper. This method may be used if a single gene modifies chromosome.
Steiner's problem in graphs is solved for the computing optimization. Population is represented as a weighted graph. Vertices of that graph represent chromosomes, edges represent the computational cost of selection criteria recurrent calculation. The GA application for identification of regression models assumes (a) gene is a regressor;
(b) chromosome is the set of regressors in single regression model (subset of all candidates);
(c) population — set of regression models (subset of all possible models); (d) selection criteria — residual sum of squares (RSS); (e) the chromosome modification by modification of one gene corresponds to the forward selection and backward elimination methods of variables (regressors) selection.
The following topics were dealt with: human/computer interfaces; texture, depth and motor perception; neural nets; fuzzy systems; learning; product/process design; simulation; robotics; visual system cybernetics; batch processes; image compression and interpretation; AI applications; fuzzy adaptive control; decision modelling; agile manufacturing; service sector; inductive algorithms; complex systems; Petri nets; real time imaging; KBS; machine recognition; requirements engineering; inspection and shop floor control; environmental decision making; medicine; supervisory control; discrete event systems; power systems; software methods; heuristic search; vision systems; database systems; information modelling; facility design and material handling; conflict resolution; emergency management; genetic algorithms; decision making and path planning; IVHS; senses approximation; intelligent user interface; robust controllers for mechanical systems; cognitive and learning systems; command and control systems; pilot associate systems; neural net applications; real time systems; mobile robot visual processes; medical applications; utility energy systems; machine recognition; computing systems design; software engineering; military applications; data analysis; stochastic processes; guided vehicles; and stability and compensation.
This research work deals with the problem formulation of control of complex organizational structures. The mechanism of functioning of such systems is described by example of a vertically integrated company (VIC). The problems of strategic and operative control of VIC are considered. The methods for solving such problems based on genetic algorithms and neural networks are suggested. A new iterative procedure for coordination of strategic and operative control goals based on the estimation of imbalance between shareholder value and net profit distributed for payment of dividends to shareholders is suggested.
The considered system is a double criterion optimization problem with complex multiparameter restrictions.
Abstract— In this paper a new multi-agent genetic algorithm for multi-objective optimization (MAGAMO) is presented. The algorithm based on the dynamical interaction of synchronized agents which are interdepended genetic algorithms (GAs) having own separate evolutions of their populations. This approach has some similarities with well known “island model” of GA. In both methods is used a migration of individuals from agents (“islands”) to the main process (“continent”). In contrast, the intelligent agents in MAGAMO are able to decompose the dimensions space to form evolutions of subpopulations (instead of distribution of initial population as in the standard “island model”). In the same time, the main (central) process is responsible for the coordination of agents only and their selection according Pareto rules (without evolution). Intelligent agents seek local suboptimal solutions for a global optimization, which will be completed in the result of the interaction of all agents. In the result of this, the amount of needed recalculating the fitness-functions can be significantly reduced. It is especially important for the multi-objective optimization related to a large-scale problem. Besides, the proposed approximating approach allows solving complex optimization problems for real big systems (like an oil company, plants, corporations, etc.).
In work the developed model of adaptive management by the vertically integrated companies based on the system approach supporting the mechanism of an operational management in a uniform cycle of strategic planning, within the limits of faster time is presented. Thus for a finding of optimum values of operating parameters special algorithms of a class of genetic algorithms are used, neural networks the example of the developed system of adaptive management for the vertically-integrated oil company is etc. presented.
In this research we analyzed the problem of solving transcendental equations' systems. More detailed is analyzed the approach of solution based on genetic algorithms because it is less examined than the one based on numerical methods. The research will be useful for different kinds of physical and mathematical calculations containing transcendental equations of high complexity. The research is available for students and graduates who are familiar with the basics of numerical methods, mathematical analysis, discrete mathematics and combinatorial algorithms.
This book constitutes the proceedings of the 16th Russian Conference on Artificial Intelligence, RCAI 2018, Moscow, Russia, in September 2018.
The 22 full papers presented along with 4 short papers in this volume were carefully reviewed and selected from 75 submissions. The conference deals with a wide range of topics, including data mining and knowledge discovery, text mining, reasoning, decision making, natural language processing, vision, intelligent robotics, multi-agent systems, machine learning, ontology engineering.
A model for organizing cargo transportation between two node stations connected by a railway line which contains a certain number of intermediate stations is considered. The movement of cargo is in one direction. Such a situation may occur, for example, if one of the node stations is located in a region which produce raw material for manufacturing industry located in another region, and there is another node station. The organization of freight traﬃc is performed by means of a number of technologies. These technologies determine the rules for taking on cargo at the initial node station, the rules of interaction between neighboring stations, as well as the rule of distribution of cargo to the ﬁnal node stations. The process of cargo transportation is followed by the set rule of control. For such a model, one must determine possible modes of cargo transportation and describe their properties. This model is described by a ﬁnite-dimensional system of diﬀerential equations with nonlocal linear restrictions. The class of the solution satisfying nonlocal linear restrictions is extremely narrow. It results in the need for the “correct” extension of solutions of a system of diﬀerential equations to a class of quasi-solutions having the distinctive feature of gaps in a countable number of points. It was possible numerically using the Runge–Kutta method of the fourth order to build these quasi-solutions and determine their rate of growth. Let us note that in the technical plan the main complexity consisted in obtaining quasi-solutions satisfying the nonlocal linear restrictions. Furthermore, we investigated the dependence of quasi-solutions and, in particular, sizes of gaps (jumps) of solutions on a number of parameters of the model characterizing a rule of control, technologies for transportation of cargo and intensity of giving of cargo on a node station.
Event logs collected by modern information and technical systems usually contain enough data for automated process models discovery. A variety of algorithms was developed for process models discovery, conformance checking, log to model alignment, comparison of process models, etc., nevertheless a quick analysis of ad-hoc selected parts of a journal still have not get a full-fledged implementation. This paper describes an ROLAP-based method of multidimensional event logs storage for process mining. The result of the analysis of the journal is visualized as directed graph representing the union of all possible event sequences, ranked by their occurrence probability. Our implementation allows the analyst to discover process models for sublogs defined by ad-hoc selection of criteria and value of occurrence probability
Existing approaches suggest that IT strategy should be a reflection of business strategy. However, actually organisations do not often follow business strategy even if it is formally declared. In these conditions, IT strategy can be viewed not as a plan, but as an organisational shared view on the role of information systems. This approach generally reflects only a top-down perspective of IT strategy. So, it can be supplemented by a strategic behaviour pattern (i.e., more or less standard response to a changes that is formed as result of previous experience) to implement bottom-up approach. Two components that can help to establish effective reaction regarding new initiatives in IT are proposed here: model of IT-related decision making, and efficiency measurement metric to estimate maturity of business processes and appropriate IT. Usage of proposed tools is demonstrated in practical cases.
Let G be a semisimple algebraic group whose decomposition into the product of simple components does not contain simple groups of type A, and P⊆G be a parabolic subgroup. Extending the results of Popov , we enumerate all triples (G, P, n) such that (a) there exists an open G-orbit on the multiple flag variety G/P × G/P × . . . × G/P (n factors), (b) the number of G-orbits on the multiple flag variety is finite.
I give the explicit formula for the (set-theoretical) system of Resultants of m+1 homogeneous polynomials in n+1 variables