Authors Yu.O. Chernyshev, N.N. Ventsov, A.A. Dolmatov
Month, Year 07, 2017 @en
Index UDC 681.3
Abstract In many cases it is difficult or impossible to obtain a priori certain training data. In this respect promising is the adaptive transfer of knowledge (transfer of learning) of the available models, context-bound areas in the projected system. This paper describes the method of fuzzy data transfer from the source task to the target. The source refers to the task with a large number of well-known (formalized) components such as objective function, system constraints, input data, etc. Target is the task with hard-formalizable parameters. It is meant that the portions of the source and target are in a context relationship. For example, the source task may consist in designing the integrated circuit on a chip of a given size. With minor geometry changes of the crystal, there is a context related problem of finishing the engineered products. The use of knowledge about the original problem while solving the task, will contribute to reducing search time. It is known that for the same optimization problems in some cases it is necessary to obtain accurate solutions, while in others it is enough to obtain approximate solutions. Under the approximate solution meant is a certain area of points, each of which describes some properties of the investigated object (process) and could be a solution to the problem, to some it is difficult to formalize the situation. It is therefore advisable to consider the procedure of the fuzzy transfer information from one subject area to another. It is shown that if the scope is vague, the procedure for the functioning of the algorithm must be modified, for example, by performing known operations on fuzzy numbers with triangular view. In practice, the transfer of fuzzy variables can be worn non-linear, it is difficult to formalize the nature. For this reason, the actual becomes the problem of finding the most appropriate transfer data from one target to another. As an example, the migrated task parameters given are the membership functions of fuzzy numbers. The advantage of the proposed approach compared to the approach J. Shell and S. Coupland, is the independence of the process of conversion from knowledge of the size of the fields definitions of the source and target tasks. The disadvantage is its dependence on the certainty of the function F describing the adequacy of the fuzzy variable transfer from one context to another.

Download PDF

Keywords Fuzzy systems; adaptation; intelligent techniques; context.
References 1. Litvinenko V.A. Adaptivnye algoritmy proektnykh operatsiy SAPR EVA [Adaptive algorithms project operations CAD EVA], IS-IT`14: Tr. Mezhdunar. kongr. po intellekt. sistemam i in-form. tekhnologiyam, p. Divnomorskoe, 2-9 sent. YuFU [IS-IT`14: proceedings of the Interna-tional. Congress on intelligent systems and technology informormation, Divnomorskoe, 2-9 September]. Moscow: Fizmatlit, 2014, Vol. 1, pp. 113-119.
2. Dey A. Understanding and Using Context, Personal and ubiquitous computing, 2001, No. 5, pp. 4-7.
3. Dourish P. What we talk about when we talk about context, Personal Ubiquitous Comput., 2004, No. 8, pp. 19-30.
4. Bettini C., Brdiczka O., Henricksen K., Indulska J., Nicklas D., Ranganathan A., Riboni D.
A survey of context modelling and reasoning techniques, Pervasive and Mobile Computing, 2010, No. 6, pp. 161-180.
5. Shell J, Coupland S. Fuzzy Transfer Learning: Methodology and Application, Preprint submit-ted to Information Sciences May 23, 2014, 27 p.
6. Borisov A.N., Krumberg O.A., Fedorov I.P. Prinyatie resheniy na osnove nechetkikh modeley: Primery ispol'zovaniya [Decision making based on fuzzy models: Examples of use]. Riga: Zinatne, 1990, 184 p.
7. Lebedev B.K., Lebedev O.B., Chernyshev Yu.O. Osnovnye zadachi sinteza topologii SBIS: Monografiya [The main tasks of the synthesis VLSI layout: Monograph]. Rostov-on-Don: RGASKhM, 2006, 92 p.
8. Borisov A.N., Krumberg O.A., Fedorov I.P. Prinyatie resheniy na osnove nechetkikh modeley: Primery ispol'zovaniya [Decision making based on fuzzy models: Examples of use]. Riga: Zinatne, 1990, 184 p.
9. Polkovnikova N.A., Kureychik V.M. Razrabotka modeli ekspertnoy sistemy na osnove nechetkoy logiki [Development of an expert system model based on fuzzy logic], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2014, No. 1 (150),
pp. 83-92.
10. Zade L.A. Fuzzy sets, Information and Control, 1965, Vol. 8, pp. 338.
11. Kureychik V.M. Osobennosti postroeniya sistem podderzhki prinyatiya resheniy [Features of decision making support system design], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2012, No. 7 (132), pp. 92-98.
12. Bershteyn L.S., Bozhenyuk A.V. Analiz ispol'zovaniya operatora implikatsii v nechetkom pravile vyvoda po analogii [Analysis of the use of the implication operator in fuzzy rule of inference by analogy], Izvestiya TRTU [Izvestiya TSURE], 2004, No. 3 (38), pp. 5-10.
13. Malyshev N.G., Bershteyn L.S., Bozhenyuk A.V. Nechetkie modeli dlya ekspertnykh sistem v SAPR [Fuzzy models for expert systems in CAD]. Moscow: Energoatomizdat, 1991, 136 p.
14. Chernyshev Yu.O., Ventsov N.N., Mukhtarov S.A. K voprosu ob intellektual'noy podderzhke protsessa dovodki SBIS [To the question of intellectual support of process of finishing VLSI], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2012, No. 7 (132), pp. 63-69.
15. Chernyshev Yu.O., Ventsov N.N., Mukhtarov S.A. Razrabotka algoritma intellektual'noy podderzhki uluchsheniya promezhutochnykh resheniy optimizatsionnykh zadach [Development of the algorithm of intellectual support improve intermediate solutions of optimization problems], Vestnik DGTU [Vestnik of DSTU], 2012, No. 5 (56), pp. 68-76.
16. Nechetkie mnozhestva v modelyakh upravleniya i iskusstvennogo intellekta [Fuzzy sets in management models and artificial intelligence], ed. by D.A. Pospelova. Moscow: Nauka. Gl. red. Fiz.-mat. lit. 1986, 321 p.
17. Zade L.A. Ponyatie lingvisticheskoy peremennoy i ego primenenie k prinyatiyu priblizhennykh resheniy [The concept of a linguistic variable and its application to making approximate deci-sions]. Moscow: Mir, 1976, 165 p.
18. Prikladnye nechetkie sistemy [Applied fuzzy systems]: translated from Japanese by K. Asai,
D. Vatada, S. Ivai and others, ed. by T. Terano, K. Asai, M. Sugeno. Moscow: Mir, 1993, 386 p.
19. Kofman A. Vvedenie v teoriyu nechetkikh mnozhestv [Introduction to the theory of fuzzy sets]: translation from French. Moscow: Radio i svyaz', 1982, 432 p.
20. Engelbrecht A. Computational intelligence: an introduction – John Wiley and Sons Ltd., 2007, 597 p.
21. Ventsov N.N. Evolyutsionnyy podkhod k modelirovaniyu raspredelitel'nykh protsessov [An evolutionary approach to modelling the distribution processes], Inzhenernyy vestnik Dona [Engineering journal of Don], 2013. Available at:

Comments are closed.