Variance Estimation For Dynamic Regression via Spectrum Thresholding

We consider the dynamic linear regression problem, where the predictor vector may vary with time. This problem can be modelled as a linear dynamical system, where the parameters that need to be learned are the variance of both the process noise and the observation noise. The classical approach to learning the variance is via the maximum likelihood estimator -- a non-convex optimization problem prone to local minima and with no finite sample complexity bounds. In this paper we study the global system operator: the operator that maps the noises vectors to the output. In particular, we obtain estimates on its spectrum, and as a result derive the {\em first known} variance estimators with sample complexity guarantees for dynamic regression problems. We evaluate the approach on synthetic and real-world benchmarks.