Asymptotically Efficient Parameter Estimation in Hidden Markov Spatio-temporal Random Fields
نویسندگان
چکیده
Estimation of the parameters of Markov random field models for spatial and temporal data arises in many applications. There are computational and statistical challenges in developing efficient estimators because of the complexity of the joint distribution of the spatio-temporal models, especially when they involve hidden states that also need to be estimated from the observations. We develop composite likelihood estimators that are analytically and computationally tractable, and show that they are asymptotically efficient under some mild correlation decay assumptions.
منابع مشابه
Model Selection for a Class of Spatio-temporal Models for Areal Data
We present a method to perform model selection based on predictive density in a class of spatio-temporal dynamic generalized linear models for areal data. These models assume a latent random field process that evolves through time with random field convolutions; the convolving fields follow proper Gaussian Markov random field processes. Parameter and latent process estimation based on Markov Ch...
متن کاملConstructing Composite Likelihoods in General Random Fields
We propose a simple estimator based on composite likelihoods for parameter learning in random field models. The estimator can be applied to all discrete graphical models such as Markov random fields and conditional random fields, including ones with higher-order energies. It is computationally efficient because it requires only inference over treestructured subgraphs of the original graph, and ...
متن کاملUnsupervised Classification of Radar Images Based on Hidden Markov Models and Generalised Mixture Estimation
Due to the enormous quantity of radar images acquired by satellites and through shuttle missions, there is an evident need for efficient automatic analysis tools. This article describes unsupervised classification of radar images in the framework of hidden Markov models and generalised mixture estimation. In particular, we show that hidden Markov chains, based on a Hilbert-Peano scan of the rad...
متن کاملConditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
We present conditional random fields , a framework for building probabilistic models to segment and label sequence data. Conditional random fields offer several advantages over hidden Markov models and stochastic grammars for such tasks, including the ability to relax strong independence assumptions made in those models. Conditional random fields also avoid a fundamental limitation of maximum e...
متن کاملAdaptive Gaussian Markov Random Fields with Applications in Human Brain Mapping
Functional magnetic resonance imaging (fMRI) has become the standard technology in human brain mapping. Analyses of the massive spatio–temporal fMRI data sets often focus on parametric or nonparametric modeling of the temporal component, while spatial smoothing is based on Gaussian kernels or random fields. A weakness of Gaussian spatial smoothing is underestimation of activation peaks or blurr...
متن کامل