Corpus ID: 174801161

Deep Compositional Spatial Models.

@article{ZammitMangion2019DeepCS,
  title={Deep Compositional Spatial Models.},
  author={Andrew Zammit‐Mangion and Tin Lok James Ng and Quan H. Vu and Maurizio Filippone},
  journal={arXiv: Methodology},
  year={2019}
}
  • Andrew Zammit‐Mangion, Tin Lok James Ng, +1 author Maurizio Filippone
  • Published 2019
  • Mathematics
  • arXiv: Methodology
  • Spatial processes with nonstationary and anisotropic covariance structure are often used when modelling, analysing and predicting complex environmental phenomena. Such processes may often be expressed as ones that have stationary and isotropic covariance structure on a warped spatial domain. However, the warping function is generally difficult to fit and not constrained to be injective, often resulting in `space-folding.' Here, we propose modelling an injective warping function through a… CONTINUE READING

    Figures and Tables from this paper.

    SXL: Spatially explicit learning of geographic processes with auxiliary tasks
    Deep Integro-Difference Equation Models for Spatio-Temporal Forecasting
    1
    Spatiotemporal Multi-Resolution Approximations for Analyzing Global Environmental Data
    Flexible spatial covariance functions
    4
    Flexible models for nonstationary dependence: Methodology and examples
    1
    Data augmentation with Möbius transformations

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 88 REFERENCES
    Hierarchical (Deep) Echo State Networks with Uncertainty Quantification for Spatio-Temporal Forecasting
    17
    Posterior inference for sparse hierarchical non-stationary models
    15
    State space models with spatial deformation
    7
    Bayesian estimation of semi‐parametric non‐stationary spatial covariance structures
    117
    Doubly Stochastic Variational Inference for Deep Gaussian Processes
    143
    Bayesian inference for non-stationary spatial covariance structure via spatial deformations
    261
    Deep Gaussian Processes
    496
    Gaussian predictive process models for large spatial data sets.
    736