Variational Marginal Particle Filters
Variational Marginal Particle Filters
复制标题
DOI：
--
复制
发表时间：
2021-09
期刊：
影响因子：
--
通讯作者：
Jinlin Lai;D. Sheldon;Justin Domke
中科院分区：
文献类型：
--
作者：
Jinlin Lai;D. Sheldon;Justin Domke
状态空间模型（SSM）的变分推理通常是困难的。最近的工作集中在衍生变分目标SSM无偏序贯蒙特卡罗估计。我们发现，边缘粒子滤波器是通过应用Rao-Blackwellization操作，牺牲了减少方差和可微性的轨迹信息，从顺序蒙特卡罗得到的。我们提出了变分边缘粒子滤波器（VMPF），这是一个可微的和可重新参数化的变分滤波目标的无偏估计的基础上的SSM。我们发现，VMPF有偏梯度提供更紧密的界限比以前的目标，和无偏的重新参数化梯度有时是有益的。
Variational inference for state space models (SSMs) is known to be hard in general. Recent works focus on deriving variational objectives for SSMs from unbiased sequential Monte Carlo estimators. We reveal that the marginal particle filter is obtained from sequential Monte Carlo by applying Rao-Blackwellization operations, which sacrifices the trajectory information for reduced variance and differentiability. We propose the variational marginal particle filter (VMPF), which is a differentiable and reparameterizable variational filtering objective for SSMs based on an unbiased estimator. We find that VMPF with biased gradients gives tighter bounds than previous objectives, and the unbiased reparameterization gradients are sometimes beneficial.