Differentiable Systematic Resampling for Variational Sequential Monte Carlo
Particle filters are a standard tool for nonlinear state estimation, but their resampling step is discrete, preventing gradient-based learning in variational sequential Monte Carlo.
ProofPaper ↗
Key points
- We introduce Differentiable Systematic Resampling (DSR), a temperature-controlled relaxation of systematic resampling, that preserves the CDF-ordered, banded structure of systematic resampling while enabling full gradient flow.
- DSR converges to exact systematic resampling as the temperature vanishes, and we prove a pointwise exponential convergence rate for the induced bias.
- Compared to optimal-transport-based differentiable resampling, DSR avoids iterative solvers and has substantially lower computational overhead.
- Experiments on stochastic dynamical systems and real-world handwriting data show that DSR achieves comparable or superior filtering and dynamics learning performance.
Sources (1)
- [1]Differentiable Systematic Resampling for Variational Sequential Monte CarloarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 03:01 PM
Particle filters are a standard tool for nonlinear state estimation, but their resampling step is discrete, preventing gradient-based learning in variational sequential Monte Carlo.
We introduce Differentiable Systematic Resampling (DSR), a temperature-controlled relaxation of systematic resampling, that preserves the CDF-ordered, banded structure of systematic resampling while enabling full gradient flow.
Extractive summary: sentences quoted from the sources.