Title :
Adaptive splitting technique for Gaussian mixture models to solve Kolmogorov Equation
Author :
Vishwajeet, Kumar ; Singla, Parveen
Author_Institution :
SUNY Univ. at Buffalo, Amherst, NY, USA
Abstract :
The accuracy and the computational complexity of a Gaussian mixture model depends upon the number of components. In a stochastic dynamical system, the number of these components must change over time to account for the change in the uncertainty over time. A new splitting technique is provided based on the minimization of Fokker Planck Kolmogorov Equation. The effect of the splitting on the other components is also discussed in the work.
Keywords :
Fokker-Planck equation; Gaussian processes; computational complexity; Fokker Planck Kolmogorov equation minimization; Gaussian mixture models; adaptive splitting technique; computational complexity; stochastic dynamical system; Accuracy; Approximation methods; Equations; Kernel; Mathematical model; Stochastic processes; Uncertainty; Nonlinear systems; Stochastic systems; Uncertain systems;
Conference_Titel :
American Control Conference (ACC), 2014
Conference_Location :
Portland, OR
Print_ISBN :
978-1-4799-3272-6
DOI :
10.1109/ACC.2014.6859240
