Abstract: This paper introduces APA (“Artificial Prion Assembly”): a pattern recognition system based on artificial prion crystalization. Specifically, the system exhibits the capability to classify patterns according to the resulting prion self- assembly simulated with cellular automata. Our approach is inspired in the biological process of proteins aggregation, known as prions, which are assembled as amyloid fibers related with neurodegenerative disorders.

Keywords: pattern recognition, prion protein aggregation, crystal growth, simulation models.

ACM Classification Keywords: F.1. Computation by abstract devices, F1.1. Models of Computation.

Link:

PRION CRYSTALIZATION MODEL AND ITS APPLICATION TO RECOGNITION PATTERN

Paula Cordero, Rafael Lahoz-Beltra? and Juan Castellanos

http://www.foibg.com/ijita/vol20/ijita20-03-p02.pdf

Abstract: The idea of Reservoir Computing has become so popular recently due to its computational efficiency and the fact that just only a linear output layer must be taught. In this paper a reservoir is proposed to be built using neo-fuzzy neurons which means that Reservoir Computing and paradigms and fuzzy logic are combined.

Keywords: reservoir computing, hybrid systems, neo-fuzzy neuron, online learning procedure, prediction.

ACM Classification Keywords: I.2.6 Learning – Connectionism and neural nets.

Link:

RESERVOIR FORECASTING NEURO-FUZZY NETWORK AND ITS LEARNING

Yevgeniy Bodyanskiy, Oleksii Tyshchenko, Iryna Pliss

http://www.foibg.com/ijita/vol20/ijita20-03-p01.pdf

Abstract: Smart Grid – is a term used to describe advanced system of energy supply, which includes various aspects from power generation to consumption, high-quality management and utility services. It also involves new digital technologies to ensure reliability, safety, and efficiency, resulting in lower costs for the provision of utility services to consumers. Since Smart Grid is intellectual and self-renewing, we can automatically solve many problems with the help of new informational technologies. The article offers the model of development of the Smart Grid based on the network architecture of the protocol G.9960. We offer the criteria of optimization and advanced data mining methods can bring new results in the field of cost-saving energy consumption and energy distribution.

Keywords: Smart Grid, protocol G.9960, recursive structure, uncertainty, functional composition, sets of rules, decision-making, knowledge database, distribution of electricity, emergency situations, consumption, data mining analysis, real time pricing.

ACM Classification Keywords: H.5 Information Systems - Information systems applications, Decision support systems, Data analytics.

Link:

METHODS AND MODELS OF THE EFFECTIVE DISTRIBUTION OF ENERGY BETWEEN CONSUMERS AND PRODUCERS

Kyzemin Oleksandr, Gurina Iryna

http://www.foibg.com/ijita/vol20/ijita20-02-p12.pdf

Abstract: Precise information and navigation monitoring in horizontal drilling is of great importance for oil exploration and deposit discovery. The cost of the drilling process depends significantly on drilling precision. The necessary drilling accuracy is provided by navigation and information-gathering instruments which need electrical power in order to operate. Small available volume, long drilling times, vibrations, and high temperatures are typical constraints for any power supply source used in downhole drilling. Downhole navigation and information-gatheringsystems become increasingly complex, with more strict demands for the parameters of their electrical power supply. Batteries and conventional power cables are the traditional and the most utilized power supply sources downhole. Both have significant disadvantages and increase the cost of drilling operations. Short lifespan of the batteries and high susceptibility to failure of the cables in the harsh downhole environment are the main motivators for further research in self-sustaining energy sources. Ideally, downhole navigation instruments and information-gathering sensors (e.g. pressure, temperature, etc.) need autonomous power supply with electrical energy produced in situ. This would improve significantly the efficacy of the entire drilling process. This article presents a review of the existing and the emerging concepts related to utilizing the available energy sources in downhole drilling. The advantages and the drawbacks of these energy-harvesting concepts are discussed and compared to conventional power sources. Progress in the development of some of these concepts is illustrated by industrial examples. Some fairly new approaches to electrical energy supply for the information exchange downhole are also presented in order to inspire non-traditional research in this field.

Keywords: Energy harvesting, Horizontal drilling processes, Oilfield Instrumentation, Measurement-While-? Drilling.

ACM Classification Keywords: A.0 General Literature - Conference proceedings; J.2 Physical Sciences and Engineering

Link:

ENERGY HARVESTING IN HORIZONTAL DRILLING PROCESSES FOR THE PURPOSE OF INFORMATION AND NAVIGATION MONITORING

Marian Dimanchev and Martin P. Mintchev

http://www.foibg.com/ijita/vol20/ijita20-02-p11.pdf

Abstract: In the article features of new portable biosensor devices of "Floratest" family for express-diagnostics of plant state in real time mode are described. The devices are developed and created in V.M.Glushkov Institute of Cybernetics of NAS of Ukraine. Today these devices are used in precision farming and environment monitoring, but new application of these devices is more challenging. Standard biosensor device can be used as part of data acquisition systems or independently, but all measurements are made in manual mode (several measurements per hour). The new one is wireless biosensor. The biosensors are integrated into sensor networks and used in precision farming and environment protection. Proposed wireless biosensors have significant differences from free-running devices. Wireless biosensors with radio transceivers are placed on test plants and data reading from set of plants is executed at the same definitive time by remote mobile computer or other mobile tools, for example, electric unmanned quadcopter. Due to it is possible to get hundreds of measurements from different parts of planted area or woodland. Implementation of proposed biosensor devices with radio channel and sensor networks on their base to agriculture and environment protection makes it possible to increase efficiency and quality of agricultural products, and spare fertilizers, water and energy resources, timely protect plants from infections and anthropogenic impact.

Keywords: wireless biosensor, precision farming, environment protection.

ACM Classification Keywords: J.3 Life and Medical Sciences - Biology and Genetics

Link:

REMOTE SMART BIOSENSORS FOR PRECISION FARMING AND ENVIRONMENT PROTECTION

Volodymyr Romanov, Dmytro Artemenko, Igor Galelyuka, Oleksandr Palagin, Yevgeniya Sarakhan

http://www.foibg.com/ijita/vol20/ijita20-02-p10.pdf

Abstract: In the paper are discussed six algorithms for non-conflict schedule obtaining in the switching nodes of type Crossbar. Comparative analysis of algorithms gives an overview of their potentiality related to the speed, required memory and performance as a function of the size N of the connection matrix T.

Keywords: Network nodes, Message switching Node traffic, Crossbar switch, Conflict elimination, Packet messages, Sparse matrix.

ACM Classification Keywords: C.2.1 Network Architecture and Design, C.4 Performance of Systems

Link:

CLASS OF ALGORITHMS FOR SYNTHESSIS OF NON-CONFLICT SCHEDULE IN COMMUNICATION NODES

Kiril Kolchakov

http://www.foibg.com/ijita/vol20/ijita20-02-p09.pdf

Abstract: The paper describes the conceptual issues of modeling cognitive processes network models at different levels. Describes possible methods construction of intelligent information processing systems based connectionism attitudes and behavior of the human mind. Proposed use of synthesis of hierarchical model based on two approaches: from neuron to concepts.

Keywords: cognitive process, network model, neural networks.

ACM Classification Keywords: I.2 ARTIFICIAL INTELLIGENCE – I.2.6 Learning – Connectionism and neural nets.

Link:

MODELING OF COGNITIVE PROCESSES BY NETWORK MODELS

Sergii Konovalenko

http://www.foibg.com/ijita/vol20/ijita20-02-p08.pdf

Abstract: Pattern recognition problem using properties of parts of a recognizable objects and relations between them is NP-hard. That is why quick recognition of an object with many elements is a difficult problem. To overcome such a difficulty the use of upper bound of steps for an algorithm solving such a problem based on logic-objective approach is suggested. The possibility of the use of Prolog language family for image recognition is investigated. An example of such a problem for which the use of these upper bounds of complexity permits to decrease the time of its solution is presented.

Keywords: pattern recognition, predicate calculus, complexity of algorithms, Prolog language.

ACM Classification Keywords: I.2.4 ARTIFICIAL INTELLIGENCE Knowledge Representation Formalisms and Methods – Predicate logic, I.5.1 PATTERN RECOGNITION Models – Deterministic, F.2.2 Nonnumerical Algorithms and Problems – Complexity of proof procedures.

Link:

EFFICIENT SIMULATION FOR PROLOG IMPLEMENTATION OF IMAGE RECOGNITION PROBLEM

Tatiana Kosovskaya, Maria Vlasova

http://www.foibg.com/ijita/vol20/ijita20-02-p07.pdf

Abstract: This paper draws parallels between speech recognition of one speaker on a limited set of words and recognition of tactile sign language. The paper also provides variant of formation of feature vectors in matrix form for both problems. It is suggested to use ellipsoidal and orthogonal compliance distances.

Keywords: speech and gesture recognition, orthogonal projectors, ellipsoidal distance, pseudoinverse, SVD – decomposition.

ACM Classification Keywords: I.2 Artificial Intelligence, I.4 Image Processing and Computer Vision, I.5 Pattern Recognition, G.1.3 Numerical Linear Algebra.

Link:

MATRIX FEATURE VECTORS IN SPEECH AND GESTURE RECOGNITION

Volodymyr Donchenko, Andrew Golik

http://www.foibg.com/ijita/vol20/ijita20-02-p06.pdf

Abstract: A mathematical notion of a computer program-function for some programming languages (in particular, for some versions of Pascal and Turbo/Visual Prolog) is introduced as a restriction of an algorithm class FP. Such a notion is useful for more adequate formalization of finite memory of computer intended for a program run. Complexity of logic-mathematical properties proof of such program-functions is investigated. A predicate language is introduced to formulate such a partial program properties. P-SPACE-completeness of elementary theories under a finite signature based on such a kind of functions with the use of halting predicate and predicates of conditional equality and conditional inequalities. This allows to use the first order language for description of logic-mathematical properties of computer program-functions by extension of the used programfunctions by means of a special additional value in the case of infinite looping of a program-function. In many cases it allows to replace traditional mathematical notions of an algorithm by the offered notion of a programfunction in the framework of mathematical informatics.

Keywords: programming languages, computer program-function, complexity classes FP, FP-SPACE.

ACM Classification Keywords: F.2.m ANALYSIS OF ALGORITHMS AND PROBLEM COMPLEXITY Miscellaneous.

Link:

ALGORITHMIC DECIDABILITY OF COMPUTER PROGRAM-FUNCTIONS LANGUAGE PROPERTIES

Nikolay Kosovskiy

http://www.foibg.com/ijita/vol20/ijita20-02-p04.pdf

Abstract: In this paper, combined exterior penalty function-conjugate gradient algorithm for a constrained optimal control problem for the coefficients of a quasilinear equation is applied. The constrained optimal control problem has been converted to one of the optimization problem using a penalty function technique. One of the approaches of building the gradient of the modified functional using the solving of the adjoint problem is investigated. A special case of the optimal control problem for the considering problem to illustrate the numerical results is investigated. The computing optimal controls are helped to identify the unknown coefficients of the quasilinear parabolic equation. The numerical results are presented.

Keywords: Optimal control, Quasilinear Parabolic Equation, Existence and Uniqueness Theorems, Adjoint system, Gradient formulae, Exterior penalty function-conjugate gradient algorithm.

ACM Classification Keywords: F.2.1 Numerical algorithm and problems,G.4 Mathematical Software

Link:

A COMBINED EXTERIOR PENALTY FUNCTION – CONJUGATE GRADIENT ALGORITHM FOR A CLASS OF CONSTRAINED OPTIMAL CONTROL QUASILINEAR PARABOLIC SYSTEMS

M. H. Farag and T. A. Talaat

http://www.foibg.com/ijita/vol20/ijita20-02-p03.pdf

Abstract: A solution for program polynomial invariant generation problem is presented. An iteration upper approximation method that was successfully applied on free algebras in this paper was adopted for polynomial ring. Set of invariants is interpreted as an ideal over polynomial ring. Relationship and intersection problems solution are proposed. Intersection of Gröbner basis is applied to solve intersection problem. Inverse obligatory is applied to solve relationship problem.

Keywords: verification theory, invariant generation, polynomial ring.

ACM Classification Keywords: D.2.4 Software/Program Verification: Assertion checkers; F.3.1 Specifying and Verifying and Reasoning about Programs: Invariants; I.2.2 Automatic Programming: Program verification; B.6.3 Design Aids : Verification; D.4.5 Reliability: Verification.

Link:

Program Invariants Generation over Polynomial Ring using Iterative Methods.

Sergii Kryvyi, Oleksandr Maksymets

http://www.foibg.com/ijita/vol20/ijita20-02-p02.pdf

Abstract: Grouping information problem appears in application in two main forms. These are the problem of recovering function, represented by its observations and the problem of classification (clusterization).It is very important for both them which are the ”representatives” of the objects under investigations: scalars, vectors or objects of other kinds. This choice is determined by the math technique can be used for handling with the “representatives”. Using the real valued vectors and Euclidean spaces correspondingly is therefore usual. Development of the technique, including SVD and Moore-Penrose? inversion on the base of special “cortege operators” for Euclidean space of Rmn type, is proposed in the article

Keywords: Feature vectors, information aggregating, generalized artificial neuronets, vector corteges, matrix corteges, linear operator between cortege spaces, Single Valued Decomposition for cortege linear operators.

ACM Classification Keywords: G.2.m. Discrete mathematics: miscellaneous,G.2.1 Combinatorics. G.3 Probability and statistics, G.1.6. Numerical analysis I.5.1.Pattern Recognition H.1.m. Models and Principles: miscellaneous:

Link:

VECTORS AND MATRIXES IN GROUPING INFORMATION PROBLEM

Donchenko V.

http://www.foibg.com/ijita/vol20/ijita20-02-p01.pdf

Abstract: The article considers innovation system modeling by the instrumentality of integrated approach. This approach consists of three main parts: differential equation system and stability analysis, agent-based modeling, «innovation funnel». We have converted our model into a flight simulator where parameters and policies can be varied on the fly and the results are displayed immediately. The model user will be able to propose and vary different strategies: number of technology brokers, share of the educational expenditures in total federal budget, operating effectiveness of technology brokers and so on.

Keywords: agent-based modeling, innovation, innovation funnel, dynamic model simulation, complex system modeling.

ACM Classification Keywords: I.2. ARTIFICIAL INTELLIGENCE: I.2.11. Distributed Artificial Intelligence – Multiagent systems.

Link:

THE DYNAMIC MODEL OF THE INNOVATION FUNNEL INDICATORS AND THEIR TRAJECTORIES

Anatoly Selyaninov, Natalia Frolova

http://www.foibg.com/ijita/vol20/ijita20-02-p05.pdf

Abstract: The basic of the hypothetical non-violent theory of interaction is provided here. Introformational method is proposed to calculate the response to non-violent influence among intellectual property rights underlying the reflex construction of intelligent systems.

Keywords: non-force interaction, introformation, reflex systems, self-organization, information theory.

ACM Classification Keywords: I.2 Artificial Intelligence – Philosophical foundations.

Link:

THEORY OF NON-VIOLENT INTERACTION

Iurii Teslia

http://www.foibg.com/ijita/vol20/ijita20-01-p05.pdf

Abstract: Отмечено, что развитие теории информатики пока не привело к единству определений отличающегося наибольшим объемом понятия информационные машины (ИМ). Цель статьи – выделение категорий актуальных компонент и разработка обобщенной схемы ИМ, овершенствование онтологии. На основе анализа взаимодействий объектов приводятся концепция структуры, основные компоненты и определение ИМ как родового понятия компьютерных и информационных систем. Показано, что для всех типов взаимодействий компонент ИМ характерным есть феномен присутствия носителей взаимодействий в виде вещественных или энергетических потоков, названных коммуникатами. В концепцию структуры ИМ введено универсальное понятие оперант: реализатор операций любого уровня, начиная от простейших - операций с коммуникатами - до самых сложных - со знаниями. По аналогии с понятиями «пиксель», воксель и др. введено и определено понятие иксели - простейшие материальные элементы, реализующие преобразование коммуникатов в информационные объекты и наоборот, а также выполняющие операции фиксации, хранения, отображения и передачи информации. По аналогии с икселем определена универсальная структура: агрегат "физический объект – информационный объект», кратко ФИОб. Множество агрегатов представлено диаграммой доменов объектов категорий R материального мира и отражений материальных объектов и ментальной деятельности. Завершенные акты взаимодействия коммуникантов, по аналогии с логической единицей работы с данными, определены как транзакции. В соответствии с ипостасью агрегата, материальной R или информационной, определены категории транзакций как , классы информационных (содержащих I-объекты), неинформационных машин (не содержащих I-объекты) и компонент ИМ.

Ключевые слова: ИНФОРМАЦИОННАЯ МАШИНА, ФУНКЦИОНАЛЬНАЯ ЦЕПЬ, ВЗАИМОДЕЙСТВИЯ, КОММУНИКАТ, ОПЕРАНТ.

ACM Classification Keywords: Theory of the Information. Philosophy and Methodology of Informatics.

Link:

ИНФОРМАЦИОННЫЕ МАШИНЫ: НЕКОТОРЫЕ КАТЕГОРИИ ФУНКЦИЙ И КОМПОНЕНТ

Мержвинский Анатолий Александрович

http://www.foibg.com/ijita/vol20/ijita20-01-p04.pdf

Abstract:The topic of the presented investigation is the automation of creative processes via one Synthetic Mata Method (SMM) and few analytic methods considered in the paper and used under SMM control. The prevention of contemporary web threats is discussed at an agent/application level. Advantages and disadvantages of synthetic data mining methods are investigated, and obstacles are revealed to their application in contemporary systems. Novel results for juxtaposing statistical vs. logical data mining methods aiming at possible evolutionary fusions are described. Recommendations are made on how to build more effective applications of classical and/or presented novel (meta) methods: SMM, KALEIDOSCOPE, FUNNEL, PUZZLE, and CONTRADICTION. The usage of ontologies is investigated with the purpose of information transfer by sense. Practical aspects of agent plications, intrusion detection, intrusion prevention, cryptography applications, multiple software and other research results are mentioned aiming to show that intelligent and classical technologies should be carefully combined in one software/hardware complex to achieve the creative goals. It is shown that all the demonstrated advantages may be combined with other known methods and technologies.

Keywords: automation of creative processes, human-machine brainstorming methods, knowledge discovery, data mining, web mining, ontology, information security systems, intrusion detection, intrusion prevention, human-centered systems, knowledge management, agent, collective evolution.

Link:

TOWARDS A NOVEL DESIGN OF CREATIVE SYSTEMS

Vladimir Jotsov

http://www.foibg.com/ijita/vol20/ijita20-01-p03.pdf

Abstract:The issues being discussed in this article are the consequences of the use of specific (journal – or article - and researcher-based) metrics (“bibliometric indices”) for assessment of the performance of scientists and research proposals.The analysis is focused on the potential of the use of such indices to operate as a mechanism for control over the production of knowledge.The methodology is based on the complexity of relationships between sciences as systems for production of knowledge and their surrounding social environment. In these interactions arise motives for control and impact over knowledge production. The effects of these motive s are expanding mechanisms for control over sciences and the knowledge they produce. The impact of the control mechanisms distorts knowledge and co-generates non-knowledge. When societies use distorted knowledge they face expansion of the so called “new risks”.On this basis “bibliometric indices” are identified as co mponents of larger (in many cases - supranational) system for control over knowledge production (sciences’ dyna mics) and as generators of distorted knowledge and unexpected and negative consequences (new risks) for societies.

Keywords: control (over sciences and knowledge); crisis of sciences; social knowledge; distorted knowledge, non-knowledge; new risks; bibliometric; academic assessment.

Link:

RISKS IN USING BIBLIOMETRIC INDICATORS FOR PERFORMANCE EVALUATION OF SCIENTISTS

Douhomir Minev

http://www.foibg.com/ijita/vol20/ijita20-01-p02.pdf

Abstract:The prevailing role of counting citations over the added scientific value evaluating distorts the scientific society. As result, the scientific work becomes a kind of business, for instance, to obtain as more citations as possible. It is important to counterbalance the role of coun ting citations by using additional qualitative criteria. The aim of this survey is to discuss an approach based on measure of “usefulness of scientific contribution” called “usc-index” and published in Markov et al, 2013. It is grounded on theory of Knowledge Market. In accordance with this, we remember main elements of this theory. After that we recall some information about Bibliometrics, Scientometrics, Informetrics and Webometrics as well as some critical analyses of journals’ metrics and quantity measures. Finally, we outline the approach for evaluation usefulness of the scientific contributions.

Keywords: Information Market, Knowledge Market, Usefulness of the Scientific Contributions

ACM Classification Keywords: A.1 Introductory and Survey

Link:

USEFULNESS OF SCIENTIFIC CONTRIBUTIONS

Krassimir Markov, KrassimiraIvanova?, Vitalii Velychko

http://www.foibg.com/ijita/vol20/ijita20-01-p01.pdf

Аннотация: Рассматриваются SDL-спецификации распределенных систем с динамическим порождением и удалением экземпляров процессов. Для них предложен метод трансляции в модифицированные цветные сети Петри - иерархические временные типизированные сети (ИЧТ- сети), в которых используется предложенная Мерлином концепция интервального времени. Естественный подход к верификации основан на использовании формальных моделей, таких, как конечные автоматы, сети Петри и их обобщения. При этом процесс анализа и верификации упрощается. Данная работа описывает SDL-системы с таймерами, которые позволяют адекватно представить значительный класс коммуникационных протоколов. Для них предложен метод трансляции в модифицированные цветные сети Петри - иерархические временные типизированные сети, в которых используются предложенная Мерлином концепция интервального времени. Алгоритм трансляции SDL-спецификаций в сетевые модели системы SDLE реализован методом трансляции в два этапа. Способ моделирования основывается на том, что в многоуровневом описании системы в SDL позиция каждого экземпляра процесса в общей иерархии системы остается неизменной, что позволяет описание системы транслировать в структуру сети, а экземпляры процесса моделировать с помощью фишек. В результате работы алгоритма создается такая сетевая модель, в которой в каждом месте будет содержать не более одной фишки, моделирующей некоторый экземпляр процесса. Таким образом, если во время функционирования системы может существовать n разных экземпляров любого процесса, то в каждом месте моделирующей его сети может содержаться не более n фишек, причем каждая фишка будет соответствовать своему экземпляру процесса. Это факт позволяет существенно повысить эффективность моделирования, так как существенно уменьшает перебор вариантов связывания переменных.

Ключевые слова: сети Петри, SDL, сетевая модель, коммуникационный протокол, экземпляр процесса.

Link:

МЕТОД ТРАНСЛЯЦИИ SDL-СПЕЦИФИКАЦИЙ С ПОМОЩЬЮ МОДИФИЦИРОВАННЫХ СЕТЕЙ ПЕТРИ ВЫСОКОГО УРОВНЯ

Анастасия Заболотная

http://www.foibg.com/ijima/vol01/ijima01-4-p09.pdf

Абстракт: В работе рассматривается подход, позволяющий проанализировать информационный потенциал системы организационного управления, с точки зрения возможности реализации режимов поддержки принятия управленческих решений. Основой анализа служит исследование основных составляющих управленческого процесса: функций-задач и функций-операций.

Ключевые слова: системы организационного управления, поддержка управленческой деятельности, источники данных, повышение эффективности управления.

Link:

ВЫБОР ИСТОЧНИКОВ ДАННЫХ ДЛЯ РЕАЛИЗАЦИИ ИНФОРМАЦИОННОЙ ПОДДЕРЖКИ УПРАВЛЕНЧЕСКОЙ ДЕЯТЕЛЬНОСТИ

Нина Баканова

http://www.foibg.com/ijima/vol01/ijima01-4-p08.pdf

Аннотация: В статье рассмотрены вопросы идентификации времени распространения пожара на основе сети TSK.

Ключевые слова: нейронные сети, экспертные оценки

ACM Classification Keywords: H.4 Information Systems Applications, J.6 Computer-aided Engineering

Link:

ИДЕНТИФИКАЦИЯ ВРЕМЕНИ РАСПРОСТРАНЕНИЯ ПОЖАРА НА ОСНОВЕ СЕТИ TSK

Александр Джулай, Артем Быченко

http://www.foibg.com/ijima/vol01/ijima01-4-p07.pdf

Аннотация: Рассмотрена задача принятия решений в условиях неопределенности с одним интервальным показателем качества сравниваемых альтернатив. Введено понятие иерархии неопределенностей в множествах предлагаемых к сравнению альтернатив, показатели качества которых описываются различными представлениями интервальных оценок, – моно интервальными и поли интервальными. Из-за типичного для задач выбора в множествах интервальных оценок пересечения интервалов задача сравнения таких, «интервальных», альтернатив может быть решена только с учетом предпочтений лица, принимающего решение. В рамках введенной иерархии предлагаются некоторые подходы к сравнению интервальных альтернатив и связанные с ними способы описания предпочтений ЛПР. Анализируется возможность введения предпочтений на базе предложенного авторами ранее метода расчета коэффициента уверенности в истинности проверяемой гипотезы о предпочтительности той или иной альтернативы. Рассматриваются возможности сравнения интервальных альтернатив и описания предпочтений посредством функций полезности, отражающих три основных вида предпочтений ЛПР, - безразличие к риску; постоянную несклонность к риску и постоянную склонность к нему. Здесь проводится также сопоставление точечных оценок, эквивалентных сравниваемым интервальным, которые рассчитываются в рамках аппарата функций полезности («детерминированные эквиваленты»), с оценками «пессимизма – оптимизма» Гурвица. Показано, что в обоих этих методах результатами сравнения выступают характерные для многократно повторяющихся ситуаций осредненные индикаторы, используемые затем для описания предпочтений, что не всегда адекватно содержанию решаемых задач. В связи с этим предложен новый метод сравнения интервальных величин и описания предпочтений, сравнение оценок и задание предпочтений в котором осуществляется на основе сопоставления «гарантированных» значений разностей показателей качества сравниваемых альтернатив, трактуемых как случайные переменные, для выбранных экспертом уровней шансов реализации проверяемой гипотезы о предпочтительности. Этот метод сравнения и описания предпочтений, который свободен от использования осредненных величин, иллюстрируется для случая полученных авторами соотношений для функций распределения вероятностей разностей двух равномерно распределенных величин, заданных на сравниваемых интервалах. Дан численный пример сравнения интервальных альтернатив, осуществляемого разными методами.

Keywords: interval alternatives, hierarchy of uncertainties, preferences, utility functions, probability distribution of difference for two random variables, methods comparing interval alternatives.

ACM Classification Keywords: H.1.2 Human information processing. G3 Distribution functions. I.2.3 Uncertainty, “fuzzy,'' and probabilistic reasoning.

Link:

ОЦЕНКА ИНТЕРВАЛЬНЫХ АЛЬТЕРНАТИВ: НЕОПРЕДЕЛЕННОСТИ И ПРЕДПОЧТЕНИЯ

Михаил Стернин, Геннадий Шепелёв

http://www.foibg.com/ijima/vol01/ijima01-4-p05.pdf

Аннотация: В работе рассматривается новый подход, ориентированный на ретроспективный анализ результативности научных проектов. Подход позволяет находить интегральные показатели оценки результативности научных проектов, используя методы группового вербального анализа решений и теорию мультимножеств. Рассмотрено применение предложенного подхода для выявления наиболее результативных научных проектов в Российском фонде фундаментальных исследований. Проведен многокритериальный анализ результатов, планируемых при подаче заявки на проект, промежуточных в ходе выполнения проекта и итоговых при завершении проекта.

Ключевые слова: групповой вербальный анализ решений, интегральный показатель оценки, результативность научного проекта, ретроспективный анализ

Link:

РЕТРОСПЕКТИВНЫЙ АНАЛИЗ РЕЗУЛЬТАТИВНОСТИ НАУЧНЫХ ПРОЕКТОВ

Алексей Петровский, Григорий Ройзензон, Александр Балышев, Игорь Тихонов

http://www.foibg.com/ijima/vol01/ijima01-4-p04.pdf

Abstract: The problem of corporations bankruptcy risk prediction is considered. Classical methods of discriminant analysis are described and analyzed. The matrix method based on fuzzy sets and new methods using fuzzy neural networks for bankruptcy risk prediction are considered, The experimental investigations of classical and new fuzzy methods for Ukrainian corporations bankruptcy risk prediction were carried out, their efficiency estimated and the best method for Ukrainian economy was determined.

Keywords: bankruptcy risk prediction, method of discriminant analysis, fuzzy neural networks

Link:

АНАЛИЗ ФИНАНСОВОГО СОСТОЯНИЯ И ПРОГНОЗИРОВАНИЕ РИСКА БАНКРОТСТВА КОРПОРАЦИЙ В УСЛОВИЯХ НЕОПРЕДЕЛЕННОСТИ

Ови Нафас Агаи Аг Гамиш, Юрий Зайченко

http://www.foibg.com/ijima/vol01/ijima01-4-p03.pdf