2016 № 2 (31)

Сontents

Bekmuratov T.F., Dadabaeva R.A. CONCEPT OF CONSTRUCTION STRATEGIC DECISION SUPPORT SYSTEM
Djumanov O.I. METHODS OF ADAPTIVE DATA PROCESSING ON THE BASIS OF MECHANISMS OF HYBRID IDENTIFICATION WITH ADJUSTMENT OF NON-STATIONARY OBJECTS MODELS PARAMETERS
Tkacheva A.A. EFFICIENT EXECUTION OF FRAGMENTED PROGRAMS USING CONTROL FLOW MEANS IN LUNA SYSTEM ON THE EXAMPLE OF DATA REDUCTION PROBLEM
Bredikhin S.V., Lyapunov V.M., Shcherbakova N.G. PARAMETERS OF NODE PAIRS OF THE CITATION NETWORK OF SCIENTIFIC PUBLICATIONS
Samigulina G.A., Massimkanova Zh.A. REVIEW OF MODERN METHODS OF SWARM INTELLIGENCE FOR COMPUTER-AIDED MOLECULAR DRUG DESIGN
Shiryayeva O.I., Denisova T.G. DEVELOPMENT OF ARTIFICIAL IMMUNE SYSTEM OF OPTIMAL CONTROL OF SULFONAMIDES THERAPEUTIC DOSES USING FUZZY LOGIC

Center for the Development of Software and Hardware-Program Complexes of Tashkent University of Information Technologies, 100084, Tashkent, Uzbekistan
*Tashkent State University of Economics, 100003, Tashkent, Uzbekistan

CONCEPT OF CONSTRUCTION STRATEGIC DECISION SUPPORT SYSTEM

UDC 684.324

The article analyzes the distinctive characteristics of the problems solved by the strategic decision support systems (DSS).Problems of the formation, structuring, storage and analytical processing of data on the changing environment of corporations, as well as the formation of recommendations for decision-making by managersare investigated. It is shown that solving these problems takes place in conditions of imperfect, incomplete information.Based on the results of the analysis the authors developed the concept of constructing strategic DSS, the main provisions of which are:

1) intellectualization, integration and consistency of interactions the functional component of the system;

2) aggregation, consistency and structure data;

3) integration of databases and knowledge;

4) the adaptability of the models of the situations and strategies of decision-making;

5) intellectuals solve problems;6) interactivity and dynamism to solving problems;

7) openness and dynamism of the functional component;

8) controllability of the accepted decisions;

9) visibility and sensitivity of the displayed information.

Explanations of these provisions.Provides a structural-functional scheme strategic intelligent DSS,constructed in accordance with the proposed concept. Described three components of the system: electronic information resources, information analysis and decision support. In the presented system uses Data Warehouse, Business Intelligence and Business Analytics, Unstructured Decision-Making and Soft Computing technologies.Discusses the information and functional relationship of the components of the system.

Key words: enterprise information system, decision support system, electronic information resource, data warehouse, data mart, data mining, business analytics.

References

1. Kenneth C. Laudon, Jane P. Laudon. Management Information Systems.Managing the Digital Firm. 12th Edition, ISBN: 9780136078463. Prentice Hall, 2012.
2. Bekmuratov T. F., Dadabaeva R. A. Globalnie korporativnie informatsionnie sistemi: predposilki, problem postroeniya i perspektivi razvitiya. [The global corporate information systems: preconditions 2C problems of construction and development prospects]. Problemiinformatiki I energetiki.[Problems of Informatics and Energetics] – Tashkent – 2015, N 1–2. P. 3–14.
3. Barsegyan A. A., Kupriyanov M. S., Kholod I. I., Tess M.D., Elizarov S. I. Analiz dannikh i protsessov; uch. posobie. [Analysis of the data and processes: tutorial] - 3-th edition. – SPb.: BKHV- Peterburg, 2009.
4. Sakharov A. A. Konseptsii postroeniya i realizatsii informatsionnikh system, orientirovannikh na analiz dannikh [Concept of construction and implementation of information systems oriented on data analysis].//SUBD [DBMS - Database management system].. – 1996.- N 4. - P. 55–70.
5. Shevelov L. V. Operativnaya analiticheskaya obrabotka dannikh: konseptsii I tekhnologii. [Online analytical processing: concepts and technologies]. http://www.olap.ru/admail.asp?Url=mailto:leonid@iname.com?subject=OLAP.
6. Prijialkovskiy V. V. Slojniy analiz dannikh bolshogo obyema; novie perspektivi komputerizatsii. [The complex analysis of large amounts of data, new prospects of computerization] //SUBD [DBMS - Database management system]. – 1996. - N 4. P. 71–83.
7. Inmon W. H. Building the Data Warehouse.- NY: John Wiley & Sons, Inc., 1992.
8. John Wang. Encyclopedia of Data Warehousing and Mining.Second Edition, 2009. IGI Global. Montclair State University, USA. Hershey • New York.Information Science reference.http://www.igi-global.com/reference.
9. Piatetsky-Fayyad U. M., Shapiro G., Smyth P. From Data Mining to Knowledge Discovery: An Overview. Advances in Knowledge Discovery and Data Mining. Edited by U. M. Fayyad, G. Piatetsky-Shapiro, P. Smyth, R. Uthurusamy. AAAI Press/The MIT Press. Menlo Park, California, Camb., Mas., London, Eng., 1996.
10. Han J., Kamber M. Data Mining.Concept and Techniques.Morgan Kaufman Publishers, 2000.
11. Parsaye K. Surveying Decision Support: New Realms of Analysis // Database Programming and Design. - 1996. - N 4.
12. Melnik O. Bizness-analitika; segonya I zavtra. [Business Analytics: Today and Tomorrow] // Journal “IntelligentEnterprise”, N 2 (212), February 2010.
13. Bekmuratov T. F. Intellektualnie informatsionnie tekhnologii: osnovnie svoystva, tipi i napravleniya razvitiya [Intelligent Information Technologies: basic properties, types and directions of development]. // TUITBulletin.Tashkent - 2008, N 1 – P.7–12.
14. Rutkovskaya D., Pilinskiy M., Rutkovskiy L. Neyronnieseti, geneticheskie algoritmi i nechetkie sistemi [Neural networks, genetic algorithms and fuzzy systems] – Moskow: Goryacayaliniya – Telekom, 2004.
15. Bekmuratov T. F., Mukhamedieva D. T. Sintez nechetko-neyronnikh modeley prinyatiya resheniy. Metodi i algoritmi sinteza nechetko-neyronnikh modeley prinyatiya resheniy [Synthesis of fuzzy-neural models of decision-making.Methods and synthesis algorithms fuzzy neural models of decision-making]Verlag “Palmarium Academic Publishing”. AVAkademikeverlagGmbH&Co. KG. Saarbruken, 2013.
16. Bekmuratov T. F., Mukhamedieva D. T. Nechetko-mnojestvennie modeli prinyatiya resheniy. Kriterii I algoritmi mnogokriterialnikh zadach prinyatiya resheniy. Verlag “Palmarium Academic Publishing”. AVAkademikeverlagGmbH&Co. KG. Saarbruken, 2015.
17. Bekmuratov T. F., Dadabaeva R. A., Mukhamedieva D. T. Prinyatiye resheniy v nechetkoy srede [Decision making in fuzzy environment]The scientific journal "Problems of Informatics" – Novosibirsk, Institute of Computational Mathematics and Mathematical Geophysics, Siberian Branch of the Russian Academy of Sciences (SB RAS), 2010 , N 1, P. 52–61.
18. Rafik A. Aliev, Oleg H. Huseynov. Decision Theory with Imperfect Information.World Scientific, 2014.http://www.worldscientific.com/worldscibooks/10.1142/9186

Bibliographic reference: Bekmuratov T. F., Dadabaeva R. A. Concept of construction strategic decision support system //journal “Problems of informatics”. 2016. № 2. P. 3-12.

Article

Djumanov O. I.

Samarkand State University, 140104, Samarkand, Uzbekistan

METHODS OF ADAPTIVE DATA PROCESSING ON THE BASIS OF MECHANISMS OF HYBRID IDENTIFICATION WITH ADJUSTMENT OF NON-STATIONARY OBJECTS MODELS PARAMETERS

UDC 658.512.011

Creating effective systems for non-stationary nature information processing to solve tasks of medical, ecological, chemical and biological researches and technological processes management is provided through the development and implementation of methods of adaptation and intellectual data analysis. Modern methods and technologies of intellectual analysis and adaptive data processing, which realized opportunities of generalization, overlapping and use of properties of neural networks, fuzzy sets and logic conclusions, are especially urgent for identification of non-stationary and poorly formalized processes in condition of limitation of prior knowledge [1,2].

Traditional approaches to constructing methods of adaptive data processing are based on widespread use of mathematical tools and software for processes modeling on the basis of statistical and dynamic identification with procedures of regulation and adjustment of models’ parameters in real time. However, the complexity of realization of adaptation mechanisms on the basis of models parameters regulation and adjustment which have been built-in in systems of data processing is caused by requirements of sufficiency prior knowledge, by absence of uncertainty and by reduction of involved volumes of computing resources [3, 4]. Consequently, the development of methods and systems to adaptive data processing, operating in conditions of limited prior information, changes of probability influences characteristics and parameters uncertainty with lower computational cost is the urgent and claimed theme of scientific research.

The present research is devoted to development of adaptive data processing methods on the basis of hybrid identification with mechanisms of parameters’ adjustment by use of neural networks (NN), which allow to receive simple, transparent and effective identifiers and approximators of non-stationary processes [5,6]. We realized methods of adaptive NN training, modified computing circuits for definition neuron’s weight, synaptic connections, functions of activation, architecture of a network, functional dependences "inputs - outputs".

In issue we investigated approaches to data processing optimization on the basis of methods for probability search with randomly enumeration, annealing, search with exclusions, stochastic modeling, forming of training sets with predominance of redundant and non-informative attributes of object [7].

The approach is offered to using models of training based on unique NN’s properties, mechanisms of adjustment of computing circuits such as network’s structural components, neurons’ weights, synaptic connections, function of activation, network architecture with specification of number of layers and neurons in layers for definition of adequate linear and nonlinear functional dependences "inputs - outputs". The modified algorithms to NN’s training, synthesized with methods of identification models parameters adjustment offer great opportunities for data processing optimization at the expense of search of the "best" parameters’ set on all NN’s components when dynamic of non-stationary process is changeable.

For development of adaptive data processing methods on the basis of NN we submit solutions of tasks of finding a suboptimum parameters’ set by reference model; forming rational training sets, database (DB), knowledgebase (KB), included properties, specific characteristics of object, rules to extracting the hidden laws of data distribution. Results are directed on ensuring stability of NN’s training with required quality of identification and approximation of casual time series (CTS) at low cost of the time for data processing.

The program complex is constructed for CTS’s identification and data processing optimization and it includes algorithms of preliminary data processing, search of rational parameters of NN’s structural components, adjustment of NN’s parameters and models of CTS description and NN’s training. The NN’s training is organized on the basis of modified computing circuits of structural components and generated training sets, which statistical parameters are adjusted on change of CTS’s dynamic properties. Thus also number of inputs, layers and neurons in layers is varied.

Developed and realized algorithms of CTS’s segmentation is based on synthesis of multistep regression model and algorithms of linear filtration, overlapping of which opportunities allow to present useful knowledge for updating NN’s training, forming suboptimum sets of identification model parameters’ adjustment for elimination of casual bursts in contours of CTS’s segments.

Functions of realized algorithms of adaptive data processing for increase of identification quality are eliminate excessive segmentation of multicomponent CTS at the expense of detecting spasmodic states of non-stationary processes, developing alternative decisions with estimating and ranking each of NN’s outputs and also at the expense of choice NN with the strongest response allowing to minimize estimate of comparison of NN’s output with characteristics of reference example.

Results of experimental researches have allowed establishing, that the essential increase of data processing systems’ efficiency is achieved at the expense of additional functional blocks, algorithms of extraction and use of data properties or features, adequate identification and approximation of CTS, algorithms of NN’s training, regulation and adjustment of NN’s parameters and models of non-stationary objects description.

Key words: non-stationary object, data processing, optimization, synthesis, statistical, dynamic, reference models, neural network, adjustment of parameters.

References

1. Miroshnik I. V., Nikiforov V. O., Fradkov A. L. Nelineynoye i adaptivnoye upravleniye slojnimi dinamicheskimi processami [Nonlinear and adaptive management of complex dynamic systems]. – S-Pb.: Nauka, 2000.
2. Rozov A. K. Obnarujeniye, klassifikasiya i osenivaniye signalov [Detection, classification and estimation of signals]. – Izdatelstvo: Politexnika, 2000.
3. Djumanov O. I. Adaptivniye algoritmi kontrolya pogreshnostey pri obuchenii neyrosetevix system obrabotki dannix nestasionarnoy prirodi [Adaptive algorithms to control errors at training of neuronetworking systems for nonstationary data processes] // “Problemi informatiki”, SO RAN, Novosibirsk. N 1(5), 2010. – pp. 61–71.
4. Yadikin I. B., Suyskiy V. M., Ovsepyan T. A. Adaptivnoye upravleniye neprerivnimi processami [Adaptive management of continuous processes]. – Moscow: Energoatomizdat, 1985.
5. Neyrosetevaya model analiza texnologicheskix vremennix ryadov v ramkax metodologii Data Mining [Neuronetworking model for analysis of time series within methodology of Data Mining] / A. M. Vulfin, A. I. Frid // Informatsionno-upravlyayushiye sistemi. N 5, 2011. – pp. 31–38.
6. Neyrosetevaya model viyavleniya I raspoznavaniya texnologicheskix situasiy v ramkax metodologii Data Mining [Neuronetworking model for revealing and recognition of technologic situations witin methodology of Data Mining] / A. M. Vulfin, A. I. Frid // Neuroinformatika-2010, Materiali XII Vserossiyskoy nauchno-texnicheskoy konferensii, MIFI. – 2010. – pp. 129–137.
7. Chipiga A. F., Voronkin R. A. Obucheniye iskustvennix neyronnix setey putyom sovmestnogo ispolzovaniya metodov lokalnoy optimizasii i geneticheskix algoritmov [Training artificial neural networks by joint using of methods of local optimization and genetic algorithms]//Izvestiya TRTU. – Tom 33, N 4. - pp. 172–174.

Bibliographic reference: Djumanov O. I. Methods of adaptive data processing on the basis of mechanisms of hybrid identification with adjustment of non-stationary objects models parameters //journal “Problems of informatics”. 2016. № 2. P. 13-20.

Article

Tkacheva A.A.

Institute of Computational Mathematics and Mathematical Geophysics SB RAS, 630090, Novosibirsk, Russia
Novosibirsk State University, 630090, Novosibirsk, Russian

EFFICIENT EXECUTION OF FRAGMENTED PROGRAMS USING CONTROL FLOW MEANS IN LUNA SYSTEM ON THE EXAMPLE OF DATA REDUCTION PROBLEM

UDC 519.685.1

With the growth of performance of parallel computers and supercomputers the complexity of system parallel programming systems also has been increased. Programmer has to think about program portability, dynamic load balance, usage all computer resources. The system of fragmented programming LuNA [3] is developed in the Institute of Computation Mathematics and Mathematical Geophysics SB RAS. The aim is to exclude parallel programming from the development of numerical models development. The fragmented programming is a parallel programming technology based on the theory of parallel program synthesis. The fragmented program (FP) is represented in declarative form as set of computation fragments (CF) and data fragments (DF). The representation doesn’t consist of the information about resources distribution and restrictions on the CF execution order other than the information dependencies. So the program can be executed on the different multicomputers including exa-flops supercomputer[15], the proper system algorithms are included into the LuNA system. The runtime system LuNA has to make decision about resources distribution, the choice what CF starts and so on in dynamics. This is the reason of the high overhead. Thus, to achieve good performance of FP execution the range LuNA program optimization transformation should be done on compilation stage. For those goals the control flow means [2,7] was developed as special kind of for-loop which executes iterations as a monolith without addition check. This substantially reduces an overhead that should be spent for next CF starting. A compiler module, that supports it, also was developed. FP execution with LuNA system was compared to FP execution with control flow means on computers with shared and distributed memory. The test demonstrated the advantages of using developed control flow means. It allows to improve on 15-5% of for-loop execution time in the LuNA system. The value of profit depends on CF amount and CF execution time inside for-loop body.

Key words: parallel programming, control flow means, fragmented programming.

References

1. Malyshkin V.E., Korneev V.D. Parallelnoe programmirovanie multicompjuterov [Parallel programming on multicomputers] // In series «Class books of NSTU», Novosibirsk, NSTU publ., 2006 – 236 p.
2. Malyshkin V.E., Perepelkin V.A., Tkacheva A.A. Control Flow Usage to Improve Performance of Fragmented Programs Execution // PaCT 2015. LNCS, vol. 9251, Springer, 2015 – pp. 86-90.
3. Malyshkin V.E., Perepelkin V.A. LuNA Fragmented Programming System, Main Functions and Peculiarities of Run-Time Subsystem. // In the Proceedings of PaCT 2013, LNCS, vol. 6873, Springer, 2011 –– pp. 53–61.
4. Bosilca, G., Bouteiller, A., Danalis, A., Herault, T., Lemarinier, P., Dongarra, J. DAGuE: A Generic Distributed DAG Engine for High Performance Computing, Proceedings of the Workshops of the 25th IEEE International Symposium on Parallel and Distributed Processing (IPDPS 2011 Workshops), IEEE, Anchorage, Alaska, USA, 1151-1158, 16-20 May, 2011.
5. Bosilca, G., Bouteiller, A., Danalis, A., Faverge, M., Haidar, A., Herault, T., Kurzak, J., Langou, J., Lemarinier, P., Ltaeif, H., Luszczek, P., YarKhan, A., Dongarra, J. "Flexible Development of Dense Linear Algebra Algorithms on Massively Parallel Architectures with DPLASMA," Proceedings of the Workshops of the 25th IEEE International Symposium on Parallel and Distributed Processing (IPDPS 2011 Workshops), IEEE, Anchorage, Alaska, USA, 1432-1441, 16-20 May, 2011.
6. Seibel P. Practical Common LISP, APRESS, 2005.
7. Tkacheva A.A. Sredstva zadaniya pryamogo upravleniya vo fragmentirovannjh programmah i ih primenenie na primere yavnogo metoda resheniya yravneniya Puassona [Control flow means of fragmented program execution on the example of the solution of Poisson equation with explicit method] // Proceedings of the Junior Researcher Conference, Institute of Computational Mathematics and Mathematical Geophysics of the SB RAS publ., Novosibirsk, 2015 – pp. 122-133.
8. Barcelona Supercomputing Center. SMP Superscalar (SMPSs) User’s Manual, Version 2.2. http://www.bsc.es/media/3576.pdf (2008)
9. D. Caromel, M. Leyton. ProActive Parallel Suite: from active objects-skeletons-components to environment and deployment. // Euro-Par 2008 Workshops – Parallel processing, 2008 – pp. 423-437.
10. Coutts, D., Loeh, A. Deterministic parallel programming with Haskell. Comput. Sci. Eng. 14 (6), 2012 – pp. 36-43.
11. Gaudiot J.-L., Deboni T., Feo J., at all. The Sisal project: real world function programming // LNCS, vol. 1808, Springer, 2001 – pp. 84-72.
12. Chao Huang, Laxmikant V. Kale Charisma. Orchestrating Migratable Parallel Objects // Proceedings of the 16th International Symposium on High Performance Distributed Computing (HPDC), 2007 – pp. 75-84.
13. Phil Miller. Productive parallel programming with Charm++ //Proceedings of the Symposium on High Performance Computing, 2015 - pp. 241-242.
14. National Research University of Novosibirsk cluster. Available at http://nusc.nsu.ru/ (accessed April 26, 2016)
15. Malyshkin V.E. Problemj parallelnoi realizacii krupnomashtabnjh chislennjh modelei na vjchislitelnjh sistemah ekzaflopsnoi proizvoditelnosti [Problems of parallel implementation of large-scale numerical models on multicomputers of exa-flops performance] // Problems of Informatics Vol. 3(28), 2015 – pp.71-82.

Bibliographic reference: Tkacheva A.A. Efficient execution of fragmented programs using control flow means in luna system on the example of data reduction problem //journal “Problems of informatics”. 2016. № 2. P. 21-29.

Article

Bredikhin S. V., Lyapunov V. M., Shcherbakova N. G.

Institute of Computational Mathematics and Mathematical Geophysics SB RAS, 630090, Novosibirsk Russia

PARAMETERS OF NODE PAIRS OF THE CITATION NETWORK OF SCIENTIFIC PUBLICATIONS

UDC 001.12+303.2

In this article methods of the analysis of the paper citation network are presented. The parameters reflecting properties of node pairs: distance, minimum cut-set, influence of common neighbors are determined and their values are calculated. The experiment showing methods of calculation of parameter values and their normalizing is fulfilled on the data retrieved from the bibliographic DB RePEc.

Key words: bibliometria, paper citation network, distance, min-cut, common neighbors, cocitation, bibliographic coupling, association, Adamic/Adar, Jaccard, Salton similarity coefficients, Katz similarity.

References

1. Bredikhin S. V., Lyapunov V. M., Shcherbakova N. G., Yurgenson A. N. Parametry “tsentral’nosti uzlov seti tsetirovaniya naucnnykh statey // Probl. inform. 2016. N 1. S. 39–57.
2. RePEc. General principles. [Electron. resource]. http://repec.org/.
3. Harari F. Teoriya grafov [Graph Theory] M.: Mir. 1973.
4. Ore O. Teoriya grafov [Graph Theory] M.: Nauka. 1980.
5. Kormen T., Leyzerson C., Rivest T. Algoritmy: postroyeniye i analiz. M.: MTSNMO. 2002. 960 p.
6. Csardi G., Nepusz T. The igraph software package for complex network research // InterJournal Complex Systems. 2006. 1695 P. [Electron. resource]. http://igraph.org/r/doc/.
7. Batagelj V., Mrvar A. Pajek Program for large network analysis [Electron. resource]. http://vlado.fmf.uni-lj.si/pub/networks/doc/pajek.pdf.
8. Network analysis (Ed. Brandes U., Erlebach T.) Berlin: Springer-Verlag. 2005. 471 p.
9. Provan J. S., Shier D. R. A paradigm for listing (s,t)-cuts in graphs // Algorithmica. 1996. V. 15. P. 351-372.
10. Ford L. R., Fulkerson D. R. Maximal flow through network // Canad. J. Math. 1956. V. 8. P. 399-404.
11. Goldberg A. V., Tarjan R. E. A new approach to the maximum-fow problem // J. of the ACM. 1988. V. 35, iss. 4. P. 921-940.
12. Kannan R., Vempala S., Vetta A. On clusterings - good, bad and spectral // Proc. of the 41st Annual IEEE Symposium on Foundations of Computer Science (FOCS'00). 2000. P. 367-378. [Electron. resource]. http://www.cc.gatech.edu/~vempala/papers/jacm-spectral.pdf.
13. Flake G. W., Tarjan R. E., Tsioutsiouliklis K. Graph clustering and minimum cut trees // [Electron. resource]. http://www.cs.princeton.edu/-kt/imath.pdf.
14. Gorke R., Hartmann T., Wagner D. Dynamic graph clustering using minimum-cut trees // J. of Graph Algorithms and Applications. 2012. V. 16, N 2. P. 411-446.
15. The link prediction problem for social networks // Proceedings of the 12th International Conference on Information and Knowledge Management, 2003. [Electron. resource]. http://www.cs.carleton.edu/faculty/dlibenno/papers/link-prediction/link.pdf.
16. Huang Z., Lin D. K. J. The time-series link prediction problem with applications in communication surveillance // INFORMS J. Computing. 2009. V. 21. P. 286-303.
17. Garfield E. Citation Indexing: A natural science literature retrieval system for the social sciences // The American Behavioral Scientist. 1964. V. 7, iss.10. P. 58-61.
18. Garner R. A computer-oriented graph theoretic analysis of citation index structure // Three Drexel information science studies. Philadelphia: Drexel Institute of Technology. 1965. P. 1-46.
19. Jarneving B. A comparison of two bibliometric methods for the mapping of the research front // Scientometrics. 2005. V. 65, iss. 2. P. 245-263.
20. Glanzel W., Czerwon H. J. A new methodological approach to bibliographic coupling and its application to the national, regional and institutional level // Scientometrics. 1996, V. 32, iss. 2. P. 195-221.
21. Garfield E. Mapping the structure of science. Citation Indexing: Its Theory and Applications in Science, Technology, and Humanities. NY: Wiley. 1979. P. 98-147. [Electron. resource]. http://www.gar-eld.library.upenn.edu/ci/chapter8.pdf.
22. Borner K., Chen C., Boyack K. W. Visualizing knowledge domains // Annual Rev. Inform. Sci. Technol. (ARIST).
23. Boyack K. W., Klavans R., Borner K. Mapping the backbone of science // Scientometrics. 2005. V. 64, iss. 3. P. 351-374.
24. Newman M. E. J. Clustering and preferential attachment in growing networks // Physical Review E. 2001. V. 64, iss. 2. 025102.
25. Newman M. E. J. Networks. An introduction. NY: Oxford University Press. 2010. 772 P.
26. Kessler M. M. Bibliographic coupling between scientific papers // Amer. Documentation. 1963. V.14, iss.1. P. 10-25.
27. Marshakova I. V. Sistema svyazey mezhdu dokumentami, postroennaya na osnove ssylok: po dannym Science Citation Index // Nauch-Techn.Inform, ser.2. 1973. N 6. S. 3–8.
28. Small H. Co-citation in the scienti-c literature: A new measure of the relationship between two documents // J. Amer. Soc. Inform. Sci 1973. V. 24, iss. 4. P. 265-269.
29. Salton G. Associative document retrieval techniques using bibliographic information // J. of the ACM. 1963. V. 10, iss. 4. P. 440-457.
30. Van Eck N. J.,Waltman L., Van den Berg J., Kaymak U. Visualizing the computational intelligence field // IEEE Computational Intelligence Magazine. 2006. V. 1, iss. 4. P. 6-10.
31. Rip A., Courtial J.-P. Co-word maps of biotechnology: An example of cognitive scientometrics // Scientometrics. 1984. V. 6, iss. 6. P. 381-400.
32. Van Eck N. J., Waltman L. How to normalize cooccurrence data? An analysis of some well-known similarity measures // J. Amer. Soc. Inform. Sci. Tech. 2009. V. 60, iss. 8. P. 1635-1651.
33. Adamic L. A., Adar E. Friends and neighbors on the web // Soc. Networks. 2003. V. 25, iss. 3. P. 211-230.
34. Katz L. A new status index derived from sociometric analysis // Psychometrika. 1953. V. 18. P. 39-43.
35. Tversky A. Features of similarity // Psychological Review. 1977. V. 84. P. 327-352.
36. Chen H., Giles C. L. ASCOS: An asymmetric network structure context similarity measure // Proc. of the 2013 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining. Niagara, ON, Canada, August 25-29. P. 442-449.
37. Acar E., Dunlavy D., Kolda T. Link prediction on evolving data using matrix and tensor factorizations // IEEE International Conference on Data Mining Workshops, 2009. Miami, Florida, December 6. P. 262-269.
38. Menger K. Zur allgemeinen Kurventheorie // Fund. Math. 1927. V. 10. P. 96-115.

Bibliographic reference: Bredikhin S.V., Lyapunov V.M., Shcherbakova N.G. Parameters of node pairs of the citation network of scientific publications //journal “Problems of informatics”. 2016. № 2. P. 30-49.

Article

Samigulina G. A., Massimkanova Zh. A.

Institute of information and computational technology, 050010, Almaty, Kazakhstan
Al-Farabi Kazakh National University, 050040, Almaty, Kazakhstan

REVIEW OF MODERN METHODS OF SWARM INTELLIGENCE FOR COMPUTER-AIDED MOLECULAR DRUG DESIGN

UDC 51-76

With rapid development of modern IT-technologies actual problem is a computer molecular design of medicines based on the approaches of artificial intelligence, which allows predicting and evaluating the impact of structural characteristics (descriptors) on their biological properties. The research of chemical compounds is connected with processing of multidimensional data sets for the solution of problem of formation an optimum set of descriptors based on certain criteria. The article provides an analytical overview of modern methods of swarm intelligence for prediction of quantitative structure-activity relationship (QSAR) of chemical substances and computer-aided molecular design of new drugs. The methods of ant and bee colonies and particle swarm optimization for solution of problem of feature selection are considered. The distinctive features of swarm intelligence algorithms are indirect exchange of information between agents, decentralizationand absence of necessity to calculate derivatives.The main software products for realization of formation of optimum set of descriptors are given.

Key words: Swarm intelligence, computer-aided molecular design, QSAR, feature selection.

References

1. Radchenko E. V., Dyаbina A. S., Palulin V. А., Zefirov N. S. Prediction of pharmacokinetic parameters for diverse drug compounds. Proceedings of the 19-th EuroQSAR. Knowledge Enabled Ligand Design. Vienna, 2012, p. 76–79.
2. Sliwoski G., Kothiwale S., Meiler J., Lowe E. W. Computational Methods in Drug Discovery. Pharmacological reviews. 2014,N 66, p. 334–395.
3. Golla S., Neely B., Whitebay E., Madihally S., Robinson R., Gasem K. Virtual design of chemical penetration enhancers for transdermal drug delivery. Chem. Biol. Drug Design. 2012, p. 478–487.
4. Chin YeeL., Chun WeiY. Current Modeling Methods Used in QSAR/QSPR. Statistical Modelling of Molecular Descriptors in QSAR/QSPR. 2012, p. 1–31.

Bibliographic reference: Samigulina G. A., Massimkanova Zh. A. Review of modern methods of swarm intelligence for computer-aided molecular drug design //journal “Problems of informatics”. 2016. № 2. P. 50-61.

Article

Shiryayeva O. I., Denisova T. G.

Institute of Information and Computing Technologies, 050010, Almaty, Kazakhstan

DEVELOPMENT OF ARTIFICIAL IMMUNE SYSTEM OF OPTIMAL CONTROL OF SULFONAMIDES THERAPEUTIC DOSES USING FUZZY LOGIC

UDC 618.5

This article presents the results of the development of immune model of organism reactions to therapeutic doses of sulfonamides based on fuzzy logic methods. There was developed an adequate mathematical model describing the body's response to drugs in the process of pyelonephritis which is used for developing optimal dynamics of the number of infected, disinfected cells depending on certain initial conditions and the drug dosage described by fuzzy sets. Also, based on the theory of fuzzy sets there were presented: scenarios of pyelonephritis extension in the body, depending on micro-organisms extension in accordance with the concepts and basics of microbiology; forms of the body infection; change of pharmacological response, depending on the sulfonamides doses based on interaction of the body and drugs in pharmacology; unfavorable drug reaction.

Today, the immune system is considered by the researchers as a source of ideas and methods of solving various problems in the analysis of biological systems with population dynamics, information processing and analysis, mathematical modeling and information security. Currently, the number of works on the development and application of artificial immune system increased rapidly.

The immune system of organism is a complex adaptive structure that effectively uses a variety of mechanisms of protection against external pathogens. The main task of the immune system is the recognition of cells (or molecules) of the body and their classification as their own or others. Detectable foreign cells serve as a signal for the activation of the protective mechanism of the appropriate type.

The immune system - is a structure in which the mechanisms of learning, memory, and associative search are implemented to solve the problems of recognition and classification. In particular, the immune system can be trained to recognize the important structures (antigenic peptides); for memorizing of already encountered structures and for the use of the laws of combinatory within gene libraries for the efficient generation of structure detectors (variable regions of antibody molecules) interacting with external antigens and the body's own cells. However, the reaction of the antigen takes place not only at the level of individual unit recognition but also at the system level through mutual recognition of lymphocyte clones in antigen-antibody reactions. Thus, the behavior of the immune system is determined by the totality of the local network interactions. Immune system attracts a great interest because of its important role in maintaining the integrity of the organism. The properties of the immune system are a remarkable example of local adaptive processes, implementing effective global responses.

Currently, under the conditions of the complex dynamics of the body's reaction to drugs the development of modeling procedures and, in the future, creating the optimal response of the body for achieving the complex of variety of purposes related to the measure of the effectiveness of the body protection, which has the plurality of evaluation criteria, is becoming one of the most important tasks of medicine. In connection with this the problems of structure creating of optimal immune system effect of drugs on the human body, where the control can be seen as a function of time, reflecting the possible impact on the disease treatment process are under a big interest. During the setting the targets for creating an optimal structure of the immune system influence of drugs on the human body in case of such complex phenomena as the processes of the body, it is quite difficult to control adequately and to construct a satisfactory measure of the quality achievement of the complex of a variety of goals that will improve the body condition, which is one of the problem for study of which it is relevant to develop specialized techniques with algorithmic programming and software.

The developing of immune model of the body reaction to drugs based on the methods of artificial immune systems should take into account the uncertainty in the parameters description, due to the nature of processes occurring in the body. Nowadays, there are various methods of uncertainty representation, also on the basis of the theory of fuzzy sets. This is conditioned by the performed analytical review of the existing methods of representation and research of fuzzy immune systems, on the basis of which there was made a conclusion about the prospects of this direction.

Today, the basic laws and mechanisms of immune system dynamics are used to evaluate and forecast the dynamics of populations of immune cells in a form of controlled mathematical models which allow exploring the protective mechanisms of the organism to the influence of external antigens. These results allow obtaining a methodology of quantitative evaluation of a therapeutic value of innovative medicines. In particular, the methodology for developing the treatment programs for complex immune diseases such as HIV-infections, improving treatment results through the use of mathematical technologies while reducing the volume of drugs; treatment of gliomas with the solving the problem of an optimal therapy strategy searching, i.e. a search of the dose and time of taking the medicine when to the final time the total number of glioma cells was minimal.

Based on the above statement the problem is composed as follows: on the basis of a mathematical model of artificial immune system of an infectious organism disease, to get the results of constructing an immune mathematical model structure simulating different algorithms of formation of special organism reactions, depending on the strategies of pyelonephritis and sulfonamides therapeutic doses based on the theory of fuzzy sets.

Key words: artificial immune system, sulfonamides, therapeutic dosage, fuzzy set.

References

1. Dasgupta D. Artificial Immune Systems and Their Applications. New York, Springer-Verlag, Inc., 1998.
2. Murray J. D. Mathematical Biology II: Spatial Models and Biomedical Applications. New-York, Springer, 2003.
3. Bolodurina I. P., Lugovskova Ju. P. Optimalnoe upravlenie dinamikoj vzaimodejstvija immunnoj sistemy cheloveka s infekcionnymi zabolevanijami [Optimal control for interaction dynamic of human immune system with infectious diseases]. Samara, Vestnik SamSU, "Mathematical modeling", 2009, N 8 (74), pp. 138–153.
4. Polat K., Sahan S., Güne S. A new method to medical diagnosis: artificial immune recognition system (AIRS) with fuzzy weighted pre-processing and application to ECG arrhythmia. Expert Systems with Applications, 31 (2), 2006, pp. 264–269.
5. Shamshirband S, Hessam S, Javidnia H, Amiribesheli M, Vahdat S, Petković D, Gani A, Kiah MLM. Tuberculosis Disease Diagnosis Using Artificial Immune Recognition System. Int. J. Med. Sci., 2014, 11(5), pp. 508–514.
6. Polat K., Kara S., Günes S. A novel approach to resource allocation mechanism in artificial immune recognition system: fuzzy resource allocation mechanism and application to diagnosis of atherosclerosis disease. Berlin: Springer-Verlag, ICARIS, 2006, pp. 244–255.
7. Shiryayeva O. I., Denisova T. G. Zadacha optimizacii terapevticheskih doz lekarstvennyh preparatov na osnove teorii nechetkih mnozhestv [Optimization problem therapeutic doses of drugs based on the theory of fuzzy sets]. International Satpaev readings "The role and place of young scientists in the implementation of the new economic policy of Kazakhstan", Almaty, 2015.
8. Soblirova Zh. H., Harina E. A. Bystryj tip acetilirovanija - vozmozhnyj marker predraspolozhennosti k zabolevanijam organov mochevoj sistemy [Fast type of acetylation - a possible marker of predisposition to diseases of the urinary system]. Nephrology and Dialysis., 1999, N 1, pp. 20–23.
9. Shiryayeva O. I., Denisova T. G. Razrabotka algoritma poluchenija optimal'nogo kolichestva sul'fanilamidov na osnove teorii iskusstvennyh immunnyh sistem [Development of receiving algorithm optimum number of sulfonamides based on artificial immune system theory]. International Scientific Conference "Theory of Functions, computer, Differential Equations and Their Applications", Kazakh National University (October 15-16), 2015.
10. Vlasov K. P. Teorija avtomaticheskogo upravlenija [Theory of control system]. – Harkov, Humanitarian center, 2007.

Bibliographic reference: Shiryayeva O. I., Denisova T. G. Development of artificial immune system of optimal control of sulfonamides therapeutic doses using fuzzy logic //journal “Problems of informatics”. 2016. № 2. P. 62-69.

Article

Main menu

You are here

2016 № 2 (31)

Сontents