350 rub
Journal Information-measuring and Control Systems №9 for 2013 г.
Article in number:
Socially significant behavior modeling on the base of super-short incomplete set of observations
Authors:
A.V. Suvorova - Reseach, SPIIRAS
Abstract:
In many fields of sociological, psychological and marketing research, we face the problem of socially significant behavior rate or frequency estimate on the base of respondents - self-reports about their behavior. We need to estimate behavior rate on the base of the responses to the questionnaire or the results of the interview. Questionnaire produces data about several behavior episodes only and about extreme intervals between them. As a result we have to analyze incomplete data.
The purpose of this paper is to provide a description of the methods for socially significant behavior modeling and estimation on the base of super-short incomplete set of observations. That model will form a base for developing algorithms for calculation of behavior rate estimates and will be a part of software system for parameters estimate.
Application of the statistical approaches is limited with the fact of very small sample size. Classical methods for small sample analysis are applicable to the samples with 10-30 items but procedures for rate estimate on the basis of the data about behavior episodes deal with 3-5 items. Time series analysis has the same restriction.
Note that these models do not provide a simple way to include additional factors such as psychosocial, demographical characteristics. One of possible ways to solve this problem is to consider a problem of socially significant behavior rate estimates in terms of probabilistic graphical models. Such formal description of the problem allows applying powerful methods and developed algorithms of the theory of Bayesian belief networks. We can use existing software to make computational simulations and apply the model to solve practical tasks.
We describe a model based on the incomplete data about time intervals between behavior episodes and propose ways of its development. Note that Bayesian belief networks has special advantages: when we consider more complex relations and as a result include more nodes to the model we just have to rearrange our model and the software does most part of calculations.
Pages: 34-37
References
- Azarov A. A., Tulup'eva T. V., Fil'chenkov A. A., Tulup'ev A. L. Veroyatnostno-relyaczionny'j podxod k predstavleniyu modeli kompleksa «informaczionnaya sistema - personal ? kritichny'e dokumenty'» // Trudy' SPIIRAN. 2012. Vy'p. 1 (20). S. 57 ? 71.
- Boks Dzh., Dzhenkins T. Analiz vremenny'x ryadov. Prognoz i upravlenie. M.: Mir. 1974. 406 s.
- Zade L. A. Rol' myagkix vy'chislenij i nechetkoj logiki v ponimanii, konstruirovanii i razvitii informaczionny'x/intellektual'ny'x sistem // Novosti Iskusstvennogo Intellekta. 2001. № 2 - 3. S. 7 - 11.
- Nevzorov V. B. Rekordy'. Matematicheskaya teoriya. M.: FAZIS, 2000. 244 s.
- Pashhenko A. E., Suvorova A. V., Tulup'eva T. V. Tulup'ev A. L. Veroyatnostny'e raspredeleniya poryadkovy'x statistik v analize sverxkorotkix nechetkix i nepolny'x vremenny'x ryadov // Trudy' SPIIRAN. 2009. Vy'p. 10. SPb.: Nauka. 2009. S. 184 - 207.
- Stepanov D. V., Musina V. F., Suvorova A. V., Tulup'ev A. L., Sirotkin A. V., Tulup'eva T. V. Funkcziya pravdopodobiya s geterogenny'mi argumentami v identifikaczii puassonovskoj modeli riskovannogo povedeniya v sluchae informaczionnogo deficzita // Trudy' SPIIRAN. 2012. Vy'p. 23. S. 157 - 184.
- Suvorova A. V., Tulup'eva T. V., Tulup'ev A. L., Sirotkin A. V., Pashhenko A. E. Veroyatnostny'e graficheskie modeli soczial'no-znachimogo povedeniya individa, uchity'vayushhie nepolnotu informaczii // Trudy' SPIIRAN. 2012. Vy'p. 22. S. 101 - 112.
- Tulup'eva T. V., Pashhenko A. E., Tulup'ev A. L., Krasnosel'skix T. V., Kazakova O. S. Modeli VICh-soczial'no-znachimogo povedeniya v kontekste psixologicheskoj zashhity' i drugix adaptivny'x stilej. SPb.: Nauka. 2008. 140 s.
- Jarushkina N. G., Afanas'eva T. V., Perfil'eva I. G. Intellektual'ny'j analiz vremenny'x ryadov: Ucheb. posobie. Ul'yanovsk: UlGTU, 2010. 320 s.
- Case K. K., Ghys P. D., Gouws E., Eaton J. W., Borquez A., Stover J., Cuchi P., Abu-Raddad L. J., Garnett G. P., Hallett T. B. Understanding the modes of transmission model of new HIV infection and its use in prevention planning // Bulletin of the World Health Organization. 2012. V. 90. № 11. P. 831 - 838A.
- GeNIe& SMILE // Decisions systems laboratory. School of Information Sciences. University of Pittsburg. http://genie.sis.pitt.edu/
- Perl J. Causality: Models, Reasoning, and Inference. Cambridge: Cambridge University Press. 2000. 400 c.
- Rubin D. B. A Small Sample Correction for Estimating Attributable Risk in Case-Control Studies // The International Journal of Biostatistics. 2010. V. 6. Iss. 1. Article 32. DOI: 10.2202/1557-4679.1252.
- Winter J. C. F., Dodou D., Wieringa P. A. Exploratory Factor Analysis With Small Sample Sizes // Multivariate Behavioral Research. 2009. 44:2. P. 147 - 181.