# Active Classification for POMDPs: a Kalman-like State Estimator

arXiv: https://arxiv.org/abs/1312.2039 · HTML: https://mlsanity.com/paper/1312.2039/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Dec 2013.

## Built on by (most cited first, 4 of 4)

- [Smoother Entropy for Active State Trajectory Estimation and Obfuscation in POMDPs](https://mlsanity.com/paper/2108.10227/) (2021)
- [Active Sampling of Multiple Sources for Sequential Estimation](https://mlsanity.com/paper/2208.05406/) (2022)
- [Entropy-Regularized Partially Observed Markov Decision Processes](https://mlsanity.com/paper/2112.12255/) (2021)
- [Learning Causal States Under Partial Observability and Perturbation](https://mlsanity.com/paper/2512.00357/) (2025)

## Abstract (arXiv)

The problem of state tracking with active observation control is considered for a system modeled by a discrete-time, finite-state Markov chain observed through conditionally Gaussian measurement vectors. The measurement model statistics are shaped by the underlying state and an exogenous control input, which influence the observations' quality. Exploiting an innovations approach, an approximate minimum mean-squared error (MMSE) filter is derived to estimate the Markov chain system state. To optimize the control strategy, the associated mean-squared error is used as an optimization criterion in a partially observable Markov decision process formulation. A stochastic dynamic programming algorithm is proposed to solve for the optimal solution. To enhance the quality of system state estimates, approximate MMSE smoothing estimators are also derived. Finally, the performance of the proposed framework is illustrated on the problem of physical activity detection in wireless body sensing networks. The power of the proposed framework lies within its ability to accommodate a broad spectrum of active classification applications including sensor management for object classification and tracking, estimation of sparse signals and radar scheduling.