Autocovariance Varieties of Moving Average Random Fields

2019/03/20 by Améndola, Carlos, Pham, Viet Son
#13P25 #14Q15 #60G60 #62F10 #62M10 #Algebraic Geometry (math.AG) #FOS: Mathematics #Statistics Theory (math.ST)

paper · doi:10.48550/arxiv.1903.08611

Abstract

We study the autocovariance functions of moving average random fields over the integer lattice ℤ^d from an algebraic perspective. These autocovariances are parametrized polynomially by the moving average coefficients, hence tracing out algebraic varieties. We derive dimension and degree of these varieties and we use their algebraic properties to obtain statistical consequences such as identifiability of model parameters. We connect the problem of parameter estimation to the algebraic invariants known as euclidean distance degree and maximum likelihood degree. Throughout, we illustrate the results with concrete examples. In our computations we use tools from commutative algebra and numerical algebraic geometry.

Related

• The Integer-valued Moving-Average Random Field
• Algebraic Statistics in Model Selection
• A crossinggram for random fields on lattices
• Central Limit Theorems for Moving Average Random Fields with Non-Random and Random Sampling On Lattices
• Commutative Algebra of Statistical Ranking