Marroquin and Ramirez (1990) have recently discovered a class of discrete stochastic cellular automata with Gibbsian invariant measures that have a non-reversible dynamic behavior. Practical applications include more powerful algorithms than the Metropolis algorithm to compute MRF models. In this paper we describe a large class of stochastic dynamical systems that has a Gibbs asymptotic distribution but does not satisfy reversibility. We characterize sufficient properties of a sub-class of stochastic differential equations in terms of the associated Fokker-Planck equation for the existence of an asymptotic probability distribution in the system of coordinates which is given. Practical implications include VLSI analog circuits to compute coupled MRF models.
| Identifer | oai:union.ndltd.org:MIT/oai:dspace.mit.edu:1721.1/6012 |
| Date | 01 December 1990 |
| Creators | Poggio, Tomaso, Girosi, Federico |
| Source Sets | M.I.T. Theses and Dissertation |
| Language | en_US |
| Detected Language | English |
| Format | 6 p., 35936 bytes, 134518 bytes, application/octet-stream, application/pdf |
| Relation | AIM-1168 |
Page generated in 0.002 seconds