Machine Learning

Congratulations to Kaspar Martens on having his paper accepted at MLCB 2023.

Generative models for multimodal data permit the identification of latent factors that may be associated with important determinants of observed data heterogeneity. Common or shared factors could be important for explaining variation across modalities whereas other factors may be private and important only for the explanation of a single modality. Multimodal Variational Autoencoders, such as MVAE and MMVAE, are a natural choice for inferring those underlying latent factors and separating shared variation from private. In this work, we investigate their capability to reliably perform this disentanglement. In particular, wehighlight a challenging problem setting where modality-specific variation dominates the shared signal. Taking a cross-modal prediction perspective, we demonstrate limitations of existing models, and propose a modification how to make them more robust to modalityspecific variation. Our findings are supported by experiments on synthetic as well as various real-world multi-omics data sets. The paper is available on PMLR.

Congratulations to group members Charles Gadd, Woojung Kim and Dominic Danks on the following papers accepted at ML4H and NeurIPS:

• mmVAE: multimorbidity clustering using Relaxed Bernoulli β-Variational Autoencoders

• Feature Allocation Approach for Multimorbidity Trajectory Modelling

My name is Christopher Yau and I am Professor of Artificial Intelligence at the University of Oxford and Health Data Research UK.

I am carrying out a survey of UK PhD students who are working in any area of data science and I need your help! We hope to get survey responses from over 300 PhD students so please help us by sparing 10-15 minutes of your time to answer some questions.

The workshop series was developed in partnership with Ovarian Cancer Action (OCA) as part of my Turing AI Fellowship.

Congratulations to PhD student Dominic Danks whose paper Derivative-Based Neural Modelling of Cumulative Distribution Functions for Survival Analysis has been accepted for presentation at AISTATS 2022.

Christopher Yau has supported Health Data Research UK (HDRUK) PhD students Fabian Falck and Haoting Zhang in the development of work that has now been published as a paper at the NeurIPS 2021 conference. The work entitled Multi-Facet Clustering Variational Autoencoders is a novel class of variational autoencoders with a hierarchy of latent variables, each with a Mixture-of-Gaussians prior, that learns multiple clusterings simultaneously, and is trained fully unsupervised and end-to-end. Chris, who directs the HDRUK PhD programme, writes about the work of the students in this blog.

Recently Christopher Yau worked with Ovarian Cancer Action UK to put together a webinar on his research for patients and the public. You can find the video on Youtube: “What is artificial intelligence and what does it mean for cancer research?”.

• Turing AI Fellowship

• OPTIMAL

• MuM-PreDICT

• Roche RPF

Suitable applicants should either have experience in developing statistical and/or machine learning methods or a background in mathematics and computational science that would enable them to learn relevant research approaches.

Congratulations to PhD student Dominic Danks whose paper BasisDeVAE: Interpretable Simultaneous Dimensionality Reduction and Feature-Level Clustering with Derivative-Based Variational Autoencoders. has been accepted for presentation at the International Conference for Machine Learning 2021.