Optimization or Architecture: How to Hack Kalman Filtering
This addresses a methodological flaw in non-linear filtering studies, potentially impacting researchers and practitioners in signal processing and machine learning, though it is incremental in optimizing an existing method.
The paper tackles the problem of unfair comparisons between non-linear architectures and the standard Kalman Filter (KF) by introducing the Optimized KF (OKF), which optimizes KF parameters similarly to neural models, making it competitive with neural models in various problems.
In non-linear filtering, it is traditional to compare non-linear architectures such as neural networks to the standard linear Kalman Filter (KF). We observe that this mixes the evaluation of two separate components: the non-linear architecture, and the parameters optimization method. In particular, the non-linear model is often optimized, whereas the reference KF model is not. We argue that both should be optimized similarly, and to that end present the Optimized KF (OKF). We demonstrate that the KF may become competitive to neural models - if optimized using OKF. This implies that experimental conclusions of certain previous studies were derived from a flawed process. The advantage of OKF over the standard KF is further studied theoretically and empirically, in a variety of problems. Conveniently, OKF can replace the KF in real-world systems by merely updating the parameters.