a link between Markov random fields and Boltzmann machines is discussed and a multivalued Boltzmann machine called vector machine model is presented.In the model
the evolution of the multivalued Boltzmann machine is implemented by using a binary Boltzmann machine. One of the advantages is the fact that the number of units of a neural network can be reduced significantly as the model represents grey images.Some examples are also presented on the application of noisy images and generation of textured images.