Dynamic Bayesian Network)

Dynamic Bayesian Network

We have developed the technology for probabilistic reasoning in the context of the static world, where each random variable has a unique fixed value. For example, when repairing a car, we always assume that the fault occurred throughout the diagnosis process is always faulty (time-independent ); our task is to deduce the state of the Car Based on the observed evidence, which remains unchanged. However, events in the real world are often related to time and are dynamic. How should we model such dynamic scenes?