Target Motion with the Metropolis-Hastings Algorithm
Target Motion with the Metropolis-Hastings Algorithm
This Demonstration shows how the Metropolis–Hastings algorithm can be used to create a random walk of target positions that corresponds to a target track as it moves through a region of interest. The overall likelihood that the target is at a location is assumed to be proportional to a likelihood function . Regions where is large correspond to regions of high target probability density and regions where is small correspond to regions of low target probability density. Specifically, the Metropolis–Hastings algorithm provides a computationally efficient method of generating a Markov chain Monte Carlo random walk that can then be interpreted as the track of a wandering target.
(x,y)
L(x,y)
L(x,y)
L(x,y)