约束和无约束多元自回归状态空间(MARSS)模型的期望最大化算法的推导
Derivation of an EM algorithm for constrained and unconstrained multivariate autoregressive state-space (MARSS) models
摘要:
本报告提出了一个期望最大化(EM)算法的约束多元自回归高斯状态空间(MARSS)模型的最大似然参数值的估计。MARSS模型可以写成:x(t)=Bx(t-1)+u+w(t),y(t)=Zx(t)+a+v(t),其中w(t)和v(t)分别是具有方差-协方差阵Q和R的多元正态误差项。MARSS模型是一类动态线性模型和向量自回归状态空间模型。Shumway和Stoffer在1982年提出了这类模型的无约束期望最大化算法,此后许多研究人员提出了特定类型的约束MARSS模型的期望最大化算法。在这份报告中,我提出了一个通用的期望最大化算法的约束MARSS模型,其中的约束是在参数矩阵(B,u,Q,Z,a,R)内的元素。约束采用vec(M)=f+Dm的形式,其中M是参数矩阵,f是固定值的列向量,D是乘数矩阵,m是估计值的列向量。这允许各种各样的约束参数矩阵形式。该演示文稿适用于时变MARSS模型,其中时间变化通过每个参数的固定(意味着未估计)f(t)和D(t)矩阵进入。该算法允许y和部分确定性系统中的缺失值,其中0出现在Q或R的对角线上。在Comprehensive R Archive Network(CRAN)的MARSS R包中提供了使用该算法估计MARSS模型的开源代码。
英文摘要:
This report presents an Expectation-Maximization (EM) algorithm for estimation of the maximum-likelihood parameter values of constrained multivariate autoregressive Gaussian state-space (MARSS) models. The MARSS model can be written: x(t)=Bx(t-1)+u+w(t), y(t)=Zx(t)+a+v(t), where w(t) and v(t) are multivariate normal error-terms with variance-covariance matrices Q and R respectively. MARSS models are a class of dynamic linear model and vector autoregressive state-space model. Shumway and Stoffer presented an unconstrained EM algorithm for this class of models in 1982, and a number of researchers have presented EM algorithms for specific types of constrained MARSS models since then. In this report, I present a general EM algorithm for constrained MARSS models, where the constraints are on the elements within the parameter matrices (B,u,Q,Z,a,R). The constraints take the form vec(M)=f+Dm, where M is the parameter matrix, f is a column vector of fixed values, D is a matrix of multipliers, and m is the column vector of estimated values. This allows a wide variety of constrained parameter matrix forms. The presentation is for a time-varying MARSS model, where time-variation enters through the fixed (meaning not estimated) f(t) and D(t) matrices for each parameter. The algorithm allows missing values in y and partially deterministic systems where 0s appear on the diagonals of Q or R. Open source code for estimating MARSS models with this algorithm is provided in the MARSS R package on the Comprehensive R Archive Network (CRAN).
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。