# Variational state filtering

March 19, 2018 — December 8, 2021

A placeholder to discuss state filtering and parameter estimation where the unobserved state is quantified by variationally-learned distributions.

Campbell et al. (2021) introduce an elegant method which also performs system identification. I would like to have time to go into more detail about this but for now I will present the key insight without adequate explanation for my own benefit. The neat trick is that the variational approximation is in a sense global, in that it all telescopes into one big variational approximation, rather than a sequence of successive approximations, each of which accumulates a greater error inside the ELBO. Intuitively this gives us more hope that we are can avoid accumulating bias at each filter step.

\[ \max _{\theta, \phi} \mathcal{L}_{t}(\theta, \phi)=\mathbb{E}_{q_{t}^{\phi}\left(x_{1: t}\right)}\left[\log \frac{p_{\theta}\left(x_{1: t}, y^{t}\right)}{q_{t}^{\phi}\left(x_{1: t}\right)}\right] \]

Our key factorization: \(q_{t}^{\phi}\left(x_{1: t}\right)=q_{t}^{\phi}\left(x_{t}\right) q_{t}^{\phi}\left(x_{t-1} \mid x_{t}\right) q_{t-1}^{\phi}\left(x_{t-2} \mid x_{t-1}\right) \ldots q_{2}^{\phi}\left(x_{1} \mid x_{2}\right)\)

True factorization: \(p_{\theta}\left(x_{t} \mid y^{t}\right) p_{\theta}\left(x_{t-1} \mid x_{t}, y^{t-1}\right) p_{\theta}\left(x_{t-2} \mid x_{t-1}, y^{t-2}\right) \cdots p_{\theta}\left(x_{1} \mid x_{2}, y^{1}\right)\)

## 1 References

*arXiv:1511.07367 [Stat]*.

*Quarterly Journal of the Royal Meteorological Society*.

*arXiv:1411.7610 [Cs, Stat]*.

*Advances in Neural Information Processing Systems 28*.

*International Journal of Approximate Reasoning*.

*Advances in Neural Information Processing Systems 24*.

*Cambridge University Engineering Department, Cambridge, England, Technical Report TR-328*.

*arXiv:1801.10395 [Stat]*.

*Bayesian Analysis*.

*Advances in Neural Information Processing Systems 30*.

*Proceedings of ICLR*.

*arXiv:1711.00799 [Stat]*.

*arXiv:1704.02798 [Cs, Stat]*.

*Advances in Neural Information Processing Systems 29*.

*Advances in Neural Information Processing Systems 27*.

*Advances in Neural Information Processing Systems 26*.

*NeuroImage*.

*Advances in Neural Information Processing Systems 28*.

*arXiv:1206.7051 [Cs, Stat]*.

*arXiv:1709.07902 [Cs, Eess, Stat]*.

*Proceedings of ICLR*.

*Mathematical and Computer Modelling of Dynamical Systems*.

*Autonomous Robots*.

*arXiv Preprint arXiv:1511.05121*.

*Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence*.

*Automatica*.

*Proceedings of The 25th International Conference on Artificial Intelligence and Statistics*.

*arXiv Preprint arXiv:1705.10306*.

*Signal Analysis and Prediction*. Applied and Numerical Harmonic Analysis.

*Proceedings of the IEEE*.

*arXiv Preprint arXiv:1603.04733*.

*arXiv Preprint arXiv:1705.09279*.

*Proceedings of ICLR*.

*Journal of Process Control*, DYCOPS-CAB 2016,.

*arXiv Preprint arXiv:1705.11140*.

*Advances in Neural Information Processing Systems 29*.

*PMLR*.

*arXiv:1802.03335 [Stat]*.

*2013 IEEE International Workshop on Machine Learning for Signal Processing (MLSP)*.

*IEEE Transactions on Automatic Control*.

*arXiv:2103.10153 [Cs, Stat]*.

*Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics*.

*Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics*.