I read many sources about kalman filter, yet no about the other approach to filtering, where canonical parametrization instead of moments parametrization is used. So I would like to learn on examples how to use this filter with information matrix and information vector (not mean and covariance from KF). Can anyone help me with this, give some examples, webpages, or parts of implementation code to be a basis to learn? Really appreciate!
Information filters are an application of a learning algorithm. One example of an information filter would be a recommendation system. On the other hand, a kalman filter is restricted to linear tracking. Its not learning per se.
If you are looking for machine learning/AI based approaches to SLAM, then you should probably look at Particle Filters.