Sparse Approximation for Gaussian Process with Derivative Observations

Ang Yang, Cheng Li, Santu Rana, Sunil Gupta, Svetha Venkatesh

Abstract

We propose a sparse Gaussian process model to approximate full Gaussian process with derivatives when a large number of function observations t and derivative observations $$t'$$ exist. By introducing a small number of inducing point m, the complexity of posterior computation can be reduced from $$\mathcal {O}((t+t')^{3})$$ to $$\mathcal {O}((t+t')m^{2})$$ . We also find the usefulness of our approach in Bayesian optimisation. Experiments demonstrate the superiority of our approach.
BibTeX
@inproceedings{yang2018sparse,
  author = {Ang Yang and Cheng Li and Santu Rana and Sunil Gupta and Svetha Venkatesh},
  title = {{Sparse Approximation for Gaussian Process with Derivative Observations}},
  booktitle = {Australasian Conference on Artificial Intelligence},
  year = {2018},
  doi = {10.1007/978-3-030-03991-2\_46},
}

Loading related content...

Figure