1. The recursive Kalman filter is known to be a ‘best’ filter in the minimum variance sense when the underlying model is correctly specified.
2. In practical applications, it may be challenging to correctly specify the stochastic model due to lack of information or computational constraints.
3. This article presents a generalized version of the Kalman filter which allows for state-vector components that are not connected in time and provides an efficient online way to describe and study the actual quality of the filter.
The article is written by experts in the field and provides a comprehensive overview of a generalized version of the Kalman filter with its precision in recursive form when the stochastic model is misspecified. The authors provide evidence for their claims through examples and references to previous research, making it clear that they have done their due diligence in researching this topic. Furthermore, they present both sides of the argument equally, noting potential risks associated with misspecifying models as well as providing solutions for how to address these issues. The only potential bias that could be identified is that some of the references used are from authors affiliated with one of the authors' institutions, which could lead to partiality in reporting results or conclusions. However, this does not appear to be an issue as all sources are cited appropriately and no promotional content was found within the article itself. All in all, this article appears to be trustworthy and reliable.