The Living Thing / Notebooks :

Particle filters and SMC

AKA “Sequential Monte Carlo” and a profusion of other simultaneous-discovery names.

A randomised generalisation of state filter models such as the Kalman Filter.

Easy to explain with an example:

A scalable particle filter in scala

EDIT: Apparently SMC is more general - it does not necessarily assume that the additional axes are assimilated in time, but can index any arbitrary dimension of your data, as long as you are approximating the right likelihood.