English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep kernel processes are a recently introduced class of deep Bayesian modelsthat have the flexibility of neural networks, but work entirely with Grammatrices. They operate by alternately sampling a Gram matrix from adistribution over positive semi-definite matrices, and applying a deterministictransformation. When the distribution is chosen to be Wishart, the model iscalled a deep Wishart process (DWP). This particular model is of interestbecause its prior is equivalent to a deep Gaussian process (DGP) prior, but atthe same time it is invariant to rotational symmetries, leading to a simplerposterior distribution. Practical inference in the DWP was made possible inrecent work ("A variational approximate posterior for the deep Wishart process"Ober and Aitchison 2021a) where the authors used a generalisation of theBartlett decomposition of the Wishart distribution as the variationalapproximate posterior. However, predictive performance in that paper was lessimpressive than one might expect, with the DWP only beating a DGP on a few ofthe UCI datasets used for comparison. In this paper, we show that furthergeneralising their distribution to allow linear combinations of rows andcolumns in the Bartlett decomposition results in better predictive performance,while incurring negligible additional computation cost.

An Improved Variational Approximate Posterior for the Deep Wishart Process