Marc Deisenroth — accepted AI-conference papers

2024

A Unifying Variational Framework for Gaussian Process Motion Planning

Lucas C. Cosier, Rares Iordan, Sicelukwanda N. T. Zwane, Giovanni Franzese, James T. Wilson, Marc Deisenroth +2

AISTATS 2024poster

To control how a robot moves, motion planning algorithms must compute paths in high-dimensional state spaces while accounting for physical constraints related to motors and joints, generating smooth and stable motions, avoiding obstacles, and preventing collisions. A motion planning algorithm must t…

Cited by 5 Source ↗PDF ↗code ↗Scholar ↗

2021

Aligning Time Series on Incomparable Spaces

Samuel Cohen, Giulia Luise, Alexander Terenin, Brandon Amos, Marc Deisenroth

AISTATS 2021poster

Dynamic time warping (DTW) is a useful method for aligning, comparing and combining time series, but it requires them to live in comparable spaces. In this work, we consider a setting in which time series live on different spaces without a sensible ground metric, causing DTW to become ill-defined. T…

Cited by 24 Source ↗PDF ↗code ↗Scholar ↗

2021

Learning Contact Dynamics using Physically Structured Neural Networks

Andreas Hochlehnert, Alexander Terenin, Steindor Saemundsson, Marc Deisenroth

AISTATS 2021poster

Learning physically structured representations of dynamical systems that include contact between different objects is an important problem for learning-based approaches in robotics. Black-box neural networks can learn to approximately represent discontinuous dynamics, but they typically require larg…

Cited by 26 Source ↗PDF ↗code ↗Scholar ↗

2021

Matérn Gaussian Processes on Graphs

Viacheslav Borovitskiy, Iskander Azangulov, Alexander Terenin, Peter Mostowsky, Marc Deisenroth, Nicolas Durrande

AISTATS 2021poster

Gaussian processes are a versatile framework for learning unknown functions in a manner that permits one to utilize prior information about their properties. Although many different Gaussian process models are readily available when the input space is Euclidean, the choice is much more limited for G…

Cited by 113 Source ↗PDF ↗Scholar ↗

2020

Efficiently sampling functions from Gaussian process posteriors

James Wilson, Viacheslav Borovitskiy, Alexander Terenin, Peter Mostowsky, Marc Deisenroth

ICML 2020poster

Gaussian processes are the gold standard for many real-world modeling problems, especially in cases where a model’s success hinges upon its ability to faithfully represent predictive uncertainty. These problems typically exist as parts of larger frameworks, wherein quantities of interest are ultimat…

Cited by 186 Source ↗PDF ↗code ↗Scholar ↗

2020

Healing Products of Gaussian Process Experts

Samuel Cohen, Rendani Mbuvha, Tshilidzi Marwala, Marc Deisenroth

ICML 2020poster

Gaussian processes (GPs) are nonparametric Bayesian models that have been applied to regression and classification problems. One of the approaches to alleviate their cubic training cost is the use of local GP experts trained on subsets of the data. In particular, product-of-expert models combine the…

Cited by 37 Source ↗PDF ↗Scholar ↗

2020

Stochastic Differential Equations with Variational Wishart Diffusions

Martin Jørgensen, Marc Deisenroth, Hugh Salimbeni

ICML 2020poster

We present a Bayesian non-parametric way of inferring stochastic differential equations for both regression tasks and continuous-time dynamical modelling. The work has high emphasis on the stochastic part of the differential equation, also known as the diffusion, and modelling it by means of Wishart…

Cited by 9 Source ↗PDF ↗code ↗Scholar ↗

2020

Variational Integrator Networks for Physically Structured Embeddings

Steindor Saemundsson, Alexander Terenin, Katja Hofmann, Marc Deisenroth

AISTATS 2020poster

Learning workable representations of dynamical systems is becoming an increasingly important problem in a number of application areas. By leveraging recent work connecting deep neural networks to systems of differential equations, we propose \emph{variational integrator networks}, a class of neural…

Cited by 79 Source ↗PDF ↗code ↗Scholar ↗

2019

Deep Gaussian Processes with Importance-Weighted Variational Inference

Hugh Salimbeni, Vincent Dutordoir, James Hensman, Marc Deisenroth

ICML 2019oral

Deep Gaussian processes (DGPs) can model complex marginal densities as well as complex mappings. Non-Gaussian marginals are essential for modelling real-world data, and can be generated from the DGP by incorporating uncorrelated variables to the model. Previous work in the DGP model has introduced n…

Cited by 60 Source ↗PDF ↗code ↗Scholar ↗

2018

Data-Efficient Reinforcement Learning with Probabilistic Model Predictive Control

Sanket Kamthe, Marc Deisenroth

AISTATS 2018poster

Trial-and-error based reinforcement learning (RL) has seen rapid advancements in recent times, especially with the advent of deep neural networks. However, the majority of autonomous RL algorithms require a large number of interactions with the environment. A large number of interactions may be impr…

Cited by 0 Source ↗PDF ↗Scholar ↗

2018

Design of Experiments for Model Discrimination Hybridising Analytical and Data-Driven Approaches

Simon Olofsson, Marc Deisenroth, Ruth Misener

ICML 2018oral

Healthcare companies must submit pharmaceutical drugs or medical device to regulatory bodies before marketing new technology. Regulatory bodies frequently require transparent and interpretable computational modelling to justify a new healthcare technology, but researchers may have several competing…

Cited by 9 Source ↗PDF ↗Scholar ↗

2018

Gaussian Process Conditional Density Estimation

Vincent Dutordoir, Hugh Salimbeni, James Hensman, Marc Deisenroth

NeurIPS 2018poster

Conditional Density Estimation (CDE) models deal with estimating conditional distributions. The conditions imposed on the distribution are the inputs of the model. CDE is a challenging task as there is a fundamental trade-off between model complexity, representational capacity and overfitting. In th…

Cited by 76 Source ↗PDF ↗code ↗Scholar ↗

2018

Maximizing acquisition functions for Bayesian optimization

James Wilson, Frank Hutter, Marc Deisenroth

NeurIPS 2018poster

Bayesian optimization is a sample-efficient approach to global optimization that relies on theoretically motivated value heuristics (acquisition functions) to guide its search process. Fully maximizing acquisition functions produces the Bayes' decision rule, but this ideal is difficult to achieve si…

Cited by 332 Source ↗PDF ↗Scholar ↗

2018

Orthogonally Decoupled Variational Gaussian Processes

Hugh Salimbeni, Ching-An Cheng, Byron Boots, Marc Deisenroth

NeurIPS 2018poster

Gaussian processes (GPs) provide a powerful non-parametric framework for reasoning over functions. Despite appealing theory, its superlinear computational and memory complexities have presented a long-standing challenge. State-of-the-art sparse variational inference methods trade modeling accuracy a…

Cited by 58 Source ↗PDF ↗code ↗Scholar ↗

2017

Doubly Stochastic Variational Inference for Deep Gaussian Processes

Hugh Salimbeni, Marc Deisenroth

NeurIPS 2017spotlight

Deep Gaussian processes (DGPs) are multi-layer generalizations of GPs, but inference in these models has proved challenging. Existing approaches to inference in DGP models assume approximate posteriors that force independence between the layers, and do not work well in practice. We present a doubly…

Cited by 541 Source ↗PDF ↗code ↗Scholar ↗

2017

Identification of Gaussian Process State Space Models

Stefanos Eleftheriadis, Tom Nicholson, Marc Deisenroth, James Hensman

NeurIPS 2017poster

The Gaussian process state space model (GPSSM) is a non-linear dynamical system, where unknown transition and/or measurement mappings are described by GPs. Most research in GPSSMs has focussed on the state estimation problem, i.e., computing a posterior of the latent state given the model. However,…

Cited by 148 Source ↗PDF ↗Scholar ↗

2015

Distributed Gaussian Processes

Marc Deisenroth, Jun Wei Ng

ICML 2015poster

To scale Gaussian processes (GPs) to large data sets we introduce the robust Bayesian Committee Machine (rBCM), a practical and scalable product-of-experts model for large-scale distributed GP regression. Unlike state-of-the-art sparse GP approximations, the rBCM is conceptually simple and does not…

Cited by 467 Source ↗PDF ↗Scholar ↗