How Deep Are Deep Gaussian Processes?

Matt Dunlop, Mark Girolami, Andrew M Stuart, Aretha L. Teckentrup

Research output: Contribution to journalArticlepeer-review


Recent research has shown the potential utility of probability distributions designed through hierarchical constructions which are conditionally Gaussian. This body of work is placed in a common framework and, through recursion, several classes of deep Gaussian processes are defined. The resulting samples have a Markovian structure with respect to the depth parameter and the effective depth of the process is interpreted in terms of the ergodicity, or non-ergodicity, of the resulting Markov chain.
Original languageEnglish
Pages (from-to)1-46
Number of pages46
JournalJournal of Machine Learning Research
Issue number54
Publication statusPublished - 9 Sep 2018


Dive into the research topics of 'How Deep Are Deep Gaussian Processes?'. Together they form a unique fingerprint.

Cite this