Abstract: The object of research in this work is homogeneous collectives of automata with the property of purposeful behavior. The subject of this study is a comparison of different designs of such machines in the implementation of the conditions of the game of placement. The aim of the study is to establish the best (or similar) structures in terms of properties in order to optimize the time and computational costs of more complex machine learning models based on the principle of reinforcement learning. In the collectives under consideration, automata perform actions in a given habitat (functioning) with varying degrees of effectiveness. The automata, in accordance with their design, react to the input signal with another action. The evaluation of the effectiveness of the machine is defined as the sum of positive signals (rewards) or negative signals (penalties) received by the machine during the considered period of time. This characteristic depends on both the declared design of the machine and the depth of its memory. It is necessary to determine the simplest designs of automata that allow achieving optimal efficiency in a given environment in the shortest possible way. The formalization of both the properties of the environment and the actions of automata, as well as the processing of the results obtained, is carried out using the apparatus of game theory. In this case, the values of the effectiveness of the functioning of the machines are represented as the cumulative amounts of winnings and losses of the slot machine players. As result of the research the designs of automata that provide a given efficiency of functioning with a minimum depth of memory (the least complex design) are presented. The result obtained makes it possible to trace the influence of the inertial qualities of automata, implemented in the form of appropriate structures, on the efficiency of functioning in a given environment, formalized in the form of a game of placement. An automaton with linear tactics and a Krylov automaton form two marginal implementations of an automaton strategy for approaching the optimum. The first is due to the high speed of changing actions, the second is due to a long stay in states close to optimal. The field of application of the results obtained is further investigation of more complex dynamic environments using the simplest designs of automata, since synchronous collectives of automata in the process of computational implementation are difficult to parallelize, which leads to a significant increase in time and computational costs with the complication of the structure of dynamic environments or with an increase in these optimization tasks.

Keywords: optimal strategy, game theory, memory depth, machine, reinforcement learning, fines, incentives, appropriate behavior, homogeneous group of machines, placement game

Abstract: The object of research is the process of choosing the optimal hardware architecture for organizing resource-intensive computing. The subject of the research is the process of solving optimization problems by genetic algorithms on GPU and CPU architectures. The influence of the choice of hardware architecture on the process of solving the optimization problem is shown: the absolute and relative dependences of the slowdown of the computing process, when choosing an irrational hardware architecture, on the number of individuals processed by the algorithm are determined. It is established that for the considered problem, the boundary of the most efficient hardware configuration can be in the range from 1000 to 5000 individuals. For this reason, it is advisable to describe the blurring of the boundary of an effective hardware configuration as a set of pairs “number of individuals — membership in a transition”. The research method is based on an analysis of the results of a computational experiment. The purpose of the experiment is to determine the dependencies of the runtime of the genetic algorithm on the GPU and CPU architectures on the number of individuals generated (chromosomes). The dependences of the minimum and maximum time of the genetic algorithm running on the GPU and CPU on the number of individuals are compared. It is shown that when solving the considered problem, the minimum and maximum time dependences of the algorithm performed on the GPU are close to a linear function; the minimum time dependences of the algorithm performed on CPU are close to a linear function, and the maximum to polynomial.

Keywords: preferred hardware architecture, intelligent system, graphics processing unit, central processing unit, adaptation, genetic algorithm, optimize calculations, evolution, multithreading, modeling

Abstract: In this paper, the author considers the problem of automatic parallelization of C programs (mainly computational) with the use of Cilk ++ directives, with the help of a limited set of which parallelism in tasks can be clearly expressed. To solve this problem, the concept of recognizing object-event models, potentially capable of parsing and transforming arbitrary texts, is formulated. This concept is a development of the theory of object-event models proposed by the author earlier, which, in the marginal formulation, are equivalent to advanced Turing machines. A general approach of the theory of object-event models is used, which asserts the possibility of describing arbitrary algorithms using these models. The technology of analysis and transformation of both structured and non-structured texts with the use of recognizing object-event models is proposed. A strategy is proposed for automatic parallelization of C programs using Cilk ++ directives based on this technology. Using the example of automatic parallelization of a simple computing program, data on acceleration and efficiency of parallelization are obtained. It is argued that the developed technology can be used as part of a program generating system for parallelizing the generated programs.

Keywords: program generation system, algorithm transformation, algorithm recognition, object-event models, Cilk, C programming language, automatic parallelizing, logical programming, task parallelism, parallel computations

Abstract: The subject of research is in the area of software development. An important task for this field is to accurately predict the timing of software development. However, the lack of standard set of tasks for which the execution time is predetermined or measured considerably complicates time management. Given these uncertainties, the author created a mathematical model of estimation of errors in predicting the timing of software development based on adaptive fuzzy system with an array of production rules, the antecedents and the consequent for which are represented by linguistic variables. The research methodology combines software engineering techniques and methods of fuzzy inference using fuzzy controllers based on production rules. The main result of the research is in building a model of estimation of errors in predicting the timing of software development based on fuzzy adaptive control system and the studies of its potential effectiveness. It is shown that the advantages of the developed model are: the possibility of obtaining error estimates based on expert judgment in the absence of statistical data; the possibility of adjusting the model during the project; the stability of the model to the one-time changes in the resulting noise values; the ability to transfer the model in a new project.

Keywords: productional logical conclusion, evaluation of programming time, software, fuzzy controller, development time program, software development, software engineering, programming costs, complexity of the software, software development project

Abstract: In this paper, we present an extended stream processors texture generation model for displaying educational content in 3D virtual worlds. The model suggests conducting image-processing tasks on stream processors in order to reduce the load on CPU. The main objective of the paper is to provide the evaluation results of the suggested extended model based on a series of tests. The extension of the model consists of using fixed pipeline features of stream processors. The obtained results of performance evaluation confirm high efficiency and veracity of the generalized mathematical and programming models for image processing. High performance can be explained by specificity problem of generating educational content for virtual words because the source data for the synthesis of images and the data area for the resultant images are in the local memory of stream processors.

Keywords: 3D virtual worlds, image processing, stream processors, educational content, vAcademia, performance evaluation, mathematical model, programming model, performance, image synthesis