|Machine learning and
A random field is a generalization of a stochastic process such that the underlying parameter need no longer be a simple real or integer valued "time", but can instead take values that are multidimensional vectors, or points on some manifold.
At its most basic, discrete case, a random field is a list of random numbers whose indices are identified with a discrete set of points in a space (for example, n-dimensional Euclidean space). When used in the natural sciences, values in a random field are often spatially correlated in one way or another. In its most basic form this might mean that adjacent values (i.e. values with adjacent indices) do not differ as much as values that are further apart. This is an example of a covariance structure, many different types of which may be modeled in a random field. More generally, the values might be defined over a continuous domain, and the random field might be thought of as a "function valued" random variable.
Definition and examples
where each is an X-valued random variable.
Several kinds of random fields exist, among them the Markov random field (MRF), Gibbs random field (GRF), conditional random field (CRF), and Gaussian random field. An MRF exhibits the Markovian property
for each choice of values , and for each , is a designated set of "neighbours" of the index point . In other words, the probability that a random variable assumes a value depends on the other random variables only through the ones that are its immediate neighbours. The probability of a random variable in an MRF is given by
where ω' is a subset of the parameter space Ω, valid for Xi.[clarification needed] It is difficult to calculate with this equation, without recourse to the relation between MRFs and GRFs proposed by Julian Besag in 1974.[clarification needed]
Random fields are of great use in studying natural processes by the Monte Carlo method, in which the random fields correspond to naturally spatially varying properties, such as soil permeability over the scale of meters, or concrete strength of the scale of centimeters. This leads to tensor random fields in which the key role is played here a Statistical Volume Element (SVE); when the SVE becomes sufficiently large, its properties become deterministic and one recovers the Representative volume element (RVE) of deterministic continuum physics. The second type of random fields that appear in continuum theories are those of dependent quantities (temperature, displacement, velocity, deformation, rotation, body and surface forces, stress, ...).
- Stochastic process
- Interacting particle system
- Stochastic cellular automata
- Vanmarcke, Erik (2010). Random Fields: Analysis and Synthesis. World Scientific Publishing Company. ISBN 978-9812563538.
- Ostoja-Starzewski, Shen and Malyarenko
- Besag, J. E. "Spatial Interaction and the Statistical Analysis of Lattice Systems", Journal of Royal Statistical Society: Series B 36, 2 (May 1974), 192-236.
- Adler, RJ & Taylor, Jonathan (2007). Random Fields and Geometry. Springer. ISBN 978-0-387-48112-8.
- Khoshnevisan (2002). Multiparameter Processes - An Introduction to Random Fields. Springer. ISBN 0-387-95459-7.
- Ostoja-Starzewski, M.; Shen, L.; Malyarenko, A. (2013), "Tensor random fields in conductivity and classical or micropolar elasticity" (PDF), Mathematics & Mechanics of Solids