# Coordinate ascent neural Kalman-MLE for state estimation

arXiv: https://arxiv.org/abs/2511.01855 · HTML: https://mlsanity.com/paper/2511.01855/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 5 of its 22 references that are in the map.
- First posted on arXiv 3 Nov 2025.

## Builds on (most cited first, 5 of 5)

- [KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics](https://mlsanity.com/paper/2107.10043/) (2021)
- [Deep State Space Models for Nonlinear System Identification](https://mlsanity.com/paper/2003.14162/) (2020)
- [Bayesian Inference and Learning in Gaussian Process State-Space Models with Particle MCMC](https://mlsanity.com/paper/1306.2861/) (2013)
- [DynaNet: Neural Kalman Dynamical Model for Motion Estimation and Prediction](https://mlsanity.com/paper/1908.03918/) (2019)
- [Sigma-Point Filtering and Smoothing Based Parameter Estimation in Nonlinear Dynamic Systems](https://mlsanity.com/paper/1504.06173/) (2015)

## Abstract (arXiv)

This paper presents a coordinate ascent algorithm to learn dynamic and measurement models in dynamic state estimation using maximum likelihood estimation in a supervised manner. In particular, the dynamic and measurement models are assumed to be Gaussian and the algorithm learns the neural network parameters that model the dynamic and measurement functions, and also the noise covariance matrices. The trained dynamic and measurement models are then used with a non-linear Kalman filter algorithm to estimate the state during the testing phase.