arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.11010econ.EMstat.CO
收藏

状态空间模型的高效变分近似

Efficient variational approximations for state space models

Rubén Loaiza-Maya, Didier Nibbering

AI总结:

针对状态空间模型现有变分贝叶斯估计不准或计算不可行的问题,本文提出适用于闭式测量密度与指数族状态转移的快速准确变分近似,并在多元 Skellam 随机波动率和时变参数 VAR 模型中验证。

摘要:

变分贝叶斯方法是状态空间模型一种具有可扩展性潜力的估计方法。然而,对于许多状态空间模型而言,现有方法要么不够准确,要么在计算上不可行。本文提出一种变分近似,适用于任何具有闭式测量密度函数、且状态转移分布属于指数族分布的模型,兼具准确性与速度。我们表明,该方法能够准确且快速地估计两类模型:一是包含四只股票高频逐笔离散价格变化的多元 Skellam 随机波动率模型,二是使用八个宏观经济变量、带随机波动率的时变参数向量自回归模型。

英文摘要:

Variational Bayes methods are a potential scalable estimation approach for state space models. However, existing methods are inaccurate or computationally infeasible for many state space models. This paper proposes a variational approximation that is accurate and fast for any model with a closed-form measurement density function and a state transition distribution within the exponential family of distributions. We show that our method can accurately and quickly estimate a multivariate Skellam stochastic volatility model with high-frequency tick-by-tick discrete price changes of four stocks, and a time-varying parameter vector autoregression with a stochastic volatility model using eight macroeconomic variables.

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。

↑