文档介绍：Application of Unscented Particle Filtering for
estimating parameters and hidden variables in gene
work
Qiang Bo, Wang Zheng-Zhi
College of Mechatronics Engineering and Automation
National University of Defense Technology,
Changsha 410073, Hunan, China
Abstract— Recent researches on estimation of parameters of gene cycle most elaborately. Recent works based on DE model are
works by differential equations generally based on generally based on Kalman Filtering (KF), the main researches
Kalman Filtering Model, it makes assumptions that the analyzed include: Wang Zi dong et al. [6]proposed a method for
system is linear. However, gene works are learning parameters work based on EM algorithm, and
obviously non-linear system, so great deviation error will predicting the expressions of genes when data were
happen. Here we present a method to estimate the parameters plete; Quach et al. [7]proposed that learning parameters
and hidden variables of gene works based on of RN based on Unscented Kalman Filtering (UKF),
Unscented Particle Filter. It makes better fitness than Kalman improving accuracy of estimation on parameters. mon
Filtering Model due to free of the premise that the model is problem exists in these works lies in that KF has the premise
linear. parison of the estimation result between that the model is linear, thus the non-linear model of