Direct Method for Estimating Parameters of Two-Segmented Piecewise Logistic Curve
Authors: Gneushev A.N., Gurchenkov A.A., Moroz I.I. | Published: 09.02.2018 |
Published in issue: #1(118)/2018 | |
DOI: 10.18698/0236-3933-2018-1-31-48 | |
Category: Informatics, Computer Engineering and Control | Chapter: Mathematical Modelling, Numerical Methods, and Program Complexes | |
Keywords: general logistic function, piecewise logistic curve, S-shape transient response, sigmoid function, mean square approximation |
In this study we examine an approach to approximation of experimental data of a two-segmented piecewise logistic curve based on the linear mean square regression. The requirement to use fixed-point arithmetic only in the estimation algorithm is the main condition for the operation in the embedded system. The paper proposes a method for parameters estimation of the piecewise logistic curve with the segment change-point coinciding with the inflection of the logistic functions. At the first stage we estimated the curve inflection point divided the samples of the experimental data into two parts. At the second stage, each part of the data is approximated independently by a generalized logistic function with the given inflection point. As a result, we suggest using a differential equation whose solution is the generalized logistic function in the mean square secondorder regression for estimating the parameters of the curve. The developed method can be used in combination with other known direct computing methods
References
[1] Balakirev V.S., Dudnikov E.G., Tsirlin A.M. Eksperimentalnoe opredelenie dinamicheskikh kharakteristik promyshlennykh obektov upravleniya [Experimental determination of the dynamic characteristics of industrial control objects]. Moscow, Energiya Publ., 1967. 230 p.
[2] Cyganek B., Socha K. Computationally efficient methods of approximations of the S-shape functions for image processing and computer graphics tasks. Image Processing & Communications, 2012, vol. 16, no. 1-2, pp. 19–28.
[3] Tisan A., Oniga S., Mic D., Buchman A. Digital implementation of the sigmoid function for FPGA circuits. Acta Technica Napocensis, Electronics and Telecommunication, 2009, vol. 50, no. 2, pp. 15–20.
[4] Pareja G.P. Fitting a logistic curve to population size data. Ph.D. Theses. Ames, Iowa, Iowa State University, 1984. 191 p.
[5] Skrobacki Z. Selected methods for the estimation of the logistic function parameters. Maintenance and Reliability, 2007, no. 3 (35), pp. 52–56.
[6] Nelder J.A. The fitting of a generalization of the logistic curve. Biometrics, 1961, vol. 17, no. 1, pp. 89–110.
[7] Jukic D., Scitovski R. Solution of the least-squares problem for logistic function. Journal of Computational and Applied Mathematics, 2003, vol. 156, no. 1, pp. 159–177.
[8] Ngunyi A., Mwita P.N., Odhiambo R.O. On the estimation and properties of logistic regression parameters. IOSR Journal of Mathematics, 2014, vol. 10, iss. 4, pp. 57–68. DOI: 10.9790/5728-10435768 Available at: http://www.iosrjournals.org/iosr-jm/papers/Vol10-issue4/Version-3/J010435768.pdf
[9] Liang J., Liang Y. Analysis and modeling for Chinas electricity demand forecasting based on a new mathematical hybrid method. Information, 2017, vol. 8, iss. 1, pp. 33–48. DOI: 10.3390/info8010033
[10] Jia Y., Li S., Tan Y., Zhao F., Hou F. Improved parametric estimation of logistic model for saturated load forecast. Power and Energy Engineering Conf. (APPEEC), Asia-Pacific Shanghai, China, 2012. DOI: 10.1109/APPEEC.2012.6307579
[11] Zhang Y., Han X., Yang G., Wang Y., Zhang L., Miao X. A novel analysis and forecast method of electricity business expanding based on seasonal adjustment. Power and Energy Engineering Conference (APPEEC), 2016, pp. 707–711. DOI: 10.1109/APPEEC.2016.7779595
[12] Cui K., Zhang L., Li J., Zhang Z., Yuan Z. Analysis and forecast of saturated load for the central city district of Tianjin. Electric Power Technologic Economics, 2008, vol. 20, no. 5, pp. 32–36.
[13] Ngufor C., Wojtusiak J. Learning from large-scale distributed health data: An approximate logistic regression approach. Proc. 30th Int. Conf. on Machine Learning, Atlanta, Georgia, USA, 2013. JMLR: W& CP, vol. 28. Available at: https://pdfs.semanticscholar.org/bf8f/e07b6304786a5efa9d313587cdc7fbed75b4.pdf
[14] Banerjee H. Estimation of parameters for logistic regression model in dose response study with a single compound or mixture of compounds. Ph.D. Theses. UC, Riverside, 2010. 77 p. Available at: http://escholarship.org/uc/item/5fv0b84p
[15] Fan G. A method to estimating the parameters of general logistic curve. Journal of Liaoning Normal University, 2009, vol. 32, no. 4, pp. 426–429.
[16] Fan G. A Method to estimating the parameters of logistic model and application. Mathematics in Economics, 2010, vol. 27, no. 1, pp. 105–110.
[17] Chu F.L. Using a logistic growth regression model to forecast the demand for tourism in Las Vegas. Tourism Management Perspectives, 2014, vol. 12, pp. 62–67. DOI: 10.1016/j.tmp.2014.08.003
[18] Pastor R., Guallar E. Use of two-segmented logistic regression to estimate changepoints in epidemiologic studies. Am. J. Epidemiol., 1998, vol. 148, no. 7, pp. 631–642.
[19] Yu J.R., Tzeng G.H., Li H.L. General fuzzy piecewise regression analysis with automatic change-point detection. Fuzzy Sets and Systems, 2001, vol. 119, iss. 2, pp. 247–257. DOI: 10.1016/S0165-0114(98)00384-4
[20] Yu J.R., Tseng F.M. Fuzzy piecewise logistic growth model for innovation diffusion: A case study of the TV industry. Int. J. Fuzzy Syst., 2016, vol. 18, iss. 3, pp. 511–522. DOI: 10.1007/s40815-015-0066-8
[21] Yu J.R., Tzeng G.H. Fuzzy multiple objective programming in an interval piecewise regression model. Int. J. Uncertain. Fuzziness Knowl.-Based Syst., 2009, vol. 17, no. 3, pp. 365–376.