350 rub
Journal Neurocomputers №3 for 2015 г.
Article in number:
Consecutive algorithms of smoothing of data
Authors:
D.A. Tarkhov - Dr.Sc. (Eng.), Professor, St.-Petersburg State Polytechnical University E-mail: dtark-hov@gmail.com
Abstract:
In article the MONS method offered B.V. Khakimovs for creation of a smoothing spline on experimental data in which serial processing of experimental points, and its some modifications is made is investigated. The numerical experiments showing are made that results of work of considered algorithms are close to the results received by exact optimization of square functionality of a mistake at much lower computing complexity. Besides, it appeared that the formula offered B.V. Khakimovs for calculation of the size of a step gives approximately the same mistake, as use of the step found optimization of functionality of a mistake at several times lower computing complexity.
Pages: 11-18
References
- KHakimov B.V. Modelirovanie korreljacionnykh zavisimostejj splajjnami na primerakh v geologii i ehkologii. M.: MGU; SPb: Neva. 2003.
- KHakimov B.V., Mikheev I.M. Nelinejjnaja model nejjrona - mnogomernyjj splajjn // Nejjrokompjutery: razrabotka, primenenie. 2012. № 7. S. 36-40.
- Tarkhov D.A. Nejjronnye seti kak sredstvo matematicheskogo modelirovanija // Nejjrokompjutery: razrabotka, primenenie. 2006. № 2. S. 3-48.
- Albert A. Regressija, psevdoinversija i rekurrentnoe ocenivanie. M.: Nauka. 1977.
- Golub Dzh., Van Loun CH. Matrichnye vychislenija. M.: Mir. 1999.