I reverted this edit of yours. I don't think "unscented" was appropriate there. Oleg Alexandrov (talk) 05:23, 10 May 2006 (UTC)
- I have been reading up on this rather extensively recently; the application is "unscented Kalman filtering," NOT "uncentered Kalman filtering." Read the "Kalman filters" page on Wikipedia for more info (I thought this would be a slam-dunk change), or check out: http://citeseer.ist.psu.edu/wan01unscented.html (with specific, explicit references to performing the Cholesky decomposition for arriving at the Sigma points used in unscented Kalman filtering), http://en.wikipedia.org/wiki/Unscented_Kalman_filter (the wikipedia page), http://www.cs.ubc.ca/~murphyk/Papers/Julier_Uhlmann_mar04.pdf (also with reference to using Cholesky decomposition as stable means of finding sigma points; page 406, footnote 6). Also, the very reference cited on the Cholesky Wikipedia page, (S. J. Julier and J.K. Uhlmann, "A new extension of the Kalman filter to nonlinear systems," in Proc. AeroSense: 11th Int. Symp. Aerospace/Defence Sensing, Simulation and Controls, 1997, pp. 182-193.), which can be found at http://www.cs.unc.edu/~welch/kalman/media/pdf/Julier1997_SPIE_KF.pdf, talks of the "unscented transform" at the heart of this (at the time) new application of the Kalman filter. The filter, henceforth, as I understand it, has been referred to as the "unscented Kalman filter." Any evidence that "uncentered" is the correct term? I am making the change again, to "unscented," and please justify a further change with citations, etc. The only Google result I found for "uncentered Kalman filter," for example, was this erroneous Wikipedia entry. --Insignificant1 06:11, 10 May 2006 (UTC)
- Insignificant1, I am sure you are right. That was your first edit, without any edit summary, so I assumed the worst. Thanks for the clarification. Oleg Alexandrov (talk) 14:57, 10 May 2006 (UTC)