PROWLER.io is headed to
NIPS 2017: The Thirty-first Annual Conference on Neural Information Processing Systems
Several of our researchers will be attending NIPS in Long Beach, California, between Dec. 4th and 9th. We're excited to present some of the ways we are advancing Principled AI.
The Annual Conference on Neural Information Processing Systems (NIPS) is the premier machine learning and computational neuroscience conference. It includes invited talks, demonstrations and oral and poster presentations of refereed papers. This year NIPS received a record-breaking 3240 submissions, of which 678 were selected for acceptance (among them 40 as 20-minute orals and 112 as 2-minute spotlights).
The PROWLER.io contingent will include CTO Dongho Kim, James Hensman, Enrique Munoz de Cote, Sergio Valcarcel Macua, Stefanos Eleftheriadis, Garrett Andersen, Tom Nicholson, Ryan-Rhys Griffiths, Mark van der Wilk and Hugh Salimbeni.
The program has been released and you can find them at the following presentations and talks:
Authors: Mark van der Wilk, Carl E Rasmussen, James Hensman (all PROWLER.io)
Convolution is a key component of machine learning tools for image recognition. Here we construct a Gaussian process model with convolutional properties and derive an effective inference procedure.
Wed Dec 6th 04:35 -- 04:50 PM @ Hall C In Probabilistic Methods, Applications
Wed Dec 6th 06:30 -- 10:30 PM @ Pacific Ballroom #196
Convolutional Gaussian Processes (pdf) in Advances in Neural Information Processing Systems 30 (NIPS 2017) pre-proceedings
Authors: Stefanos Eleftheriadis, Thomas F.W. Nicholson, Marc Peter Deisenroth, James Hensman (all PROWLER.io)
The Gaussian process state space model (GPSSM) is a powerful model of nonlinear dynamical systems. We present a practical approach to inference in these systems using a recurrent neural network as a recognition model.
Tue Dec 5th 06:30 -- 10:30 PM @ Pacific Ballroom #190
Authors: Hugh Salimbeni (PROWLER.io and Imperial College), Marc Deisenroth (PROWLER.io and Imperial College)
We present an algorithm for inference in deep Gaussian processes. The method requires fewer factorisation assumptions than previous methods and works very well in practice.
Wed Dec 6th 05:25 -- 05:30 PM @ Hall C In Probabilistic Methods, Applications
Wed Dec 6th 06:30 -- 10:30 PM @ Pacific Ballroom #195
Authors: Garrett Andersen (PROWLER.io), George Konidaris (Brown)
We introduce an online active exploration algorithm for data-efficiently learning an abstract symbolic model of an environment. Our agent uses a Bayesian symbolic model to guide its exploration towards regions of the state space that the model is uncertain about.
Wed Dec 6th 06:30 -- 10:30 PM @ Pacific Ballroom #7
Ryan-Rhys Griffiths (PROWLER.io), José Miguel Hernández-Lobato (Cambridge)
We apply Constrained Bayesian Optimization to yield an improved generative model of novel drug compounds.
Friday Dec 8th, 2017 at 10:10 Room 102-C, Long Beach Convention Center as part of the Workshop: Machine Learning for Molecules and Materials Fri Dec 8th 08:00 AM -- 06:30 PM @ S4
For abstracts and more papers from PROWLER.io, please visit our research page.