Credits: Mike Mackenzie

AI4Science Colloquium

Knowledge Shared = Knowledge2

Previous Colloquium

Density functional theory and machine learning: The race to find the next best approximation

Date: Tuesday 11 June 2024 at 11:00-1200 CET
Place: Science Park 904, Room A1.28

Kieron Burke

Speaker: Kieron Burke, University of California at Irive, USA.

Abstract:

At least 50,000 papers each year report the results of Kohn-Sham density functional calculations for materials and molecular properties. This is a huge worldwide effort, growing rapidly with computer power and powerful machine-learning algorithms to search for desired properties. But all these calculations are limited by the accuracy and generality of our current approximations. I will discuss the race to use machine learning to find the next best approximate functional, and the potential impact.

Bio:

Professor Kieron Burke in a distinguished professor in the Chemistry department and the Physics department at UC Irvine. Prof. Burke is well-known for his fundamental work on Density Functional Theory. The “B” in the popular “PBE”-functional stands for Burke, so several of us have probably used this DFT functional that he co-developed. He has worked extensively on the DFT formalism, extensions to new areas, and applications in chemistry, materials science, matter under extreme conditions, and so forth. Recently, he has also focused on errors in DFT and machine learned density functionals.

Previous colloquia

Click on the items below for further details.

14 March 2023 › Aaditya Ramdas
Conformal prediction under distribution shift
Read More ›


25 October 2022 › Fredrik Lindsten
Deep Gaussian Markov Random Fields
Read More ›


11 October 2022 › Guy Wolf
Geometry-based Data Exploration
Read More ›


5 July 2022 › Michele Ceriotti
A unified theory of atom cloud representations for machine learning
Read More ›


21 June 2022 › Peter Grünwald
E is the New P: Evidence and Optional Continuation
Read More ›


7 June 2022 › Wujie Wang
Generative Coarse-Graining of Molecular Conformations
Read More ›


24 May 2022 › Francesca Grisoni
Harnessing artificial intelligence for de novo drug design
Read More ›


26 April 2022 › Maximilian Dax
Real-Time Gravitational Wave Science with Neural Posterior Estimation
Read More ›


14 April 2022 › Gabriel Vivo-Truyols
On the use of Bayesian statistics for chromatography and mass spectrometry: dealing with big data
Read More ›


29 March 2022 › Anna Scaife
Catalogue curation, likelihood misspecification and dataset shift: challenges for Bayesian deep-learning in radio astronomy
Read More ›


15 March 2022 › Rajesh Ranganath
Interpretability and Out of Distribution Generalization in Deep Predictive Models
Read More ›


1 March 2022 › Martin van Hecke
Machine Learning of Combinatiorial Rules in Mechanical Metamaterials
Read More ›


7 February 2022 › Jan-Matthis Lückmann
Simulation-Based Inference for Neuroscience and Beyond
Read More ›


18 January 2022 › Andrew Ferguson
Ultra-fast molecular simulators and data-driven protein design
Read More ›


21 December 2021 › Atılım Güneş Baydin
Probabilistic Programming in Scientific Simulators
Read More ›


23 November 2021 › Chris Rackauckas
The Interplay of Science-Guided AI and Differentiable Simulation
Read More ›


9 November 2021 › Jan-Willem van de Meent
Compositional Inference in Probabilistic Programs
Read More ›


31 August 2021 › Saer Samanipour
Machine Learning and High Resolution Mass Spectrometry
Read More ›


22 June 2021 › Pratyush Tiwari
Can artificial intelligence help understand and predict molecular dynamics?
Read More ›


25 May 2021 › Alexander Tkatchenko
On Electrons and Machine Learning Force Fields
Read More ›


11 May 2021 › Chandler Squires
Predicting Gene Expression across Cell Types and Drugs
Read More ›


13 April 2021 › Ferry Hooft
Discovering Collective Variables of Molecular Transitions via Genetic Algorithms and Neural Networks
Read More ›


30 March 2021 › Benjamin Miller
Constrained marginal likelihood-to-evidence ratio estimation
Read More ›


2 March 2021 › Jakub Tomczak
All that glitters is not Deep Learning in Life Sciences (but sometimes it is!)
Read More ›


16 February 2021 › Eliu Huerta Escudero
Towards Accelerated, Reproducible, Physics-informed AI-driven Discovery
Read More ›


2 February 2021 › Mario Geiger
e3nn: Euclidean symmetry for neural networks
Read More ›


19 January 2021 › Dim Coumou
Machine Learning in climate science: Finding causal connections and improving seasonal forecasts
Read More ›


24 November 2020 › Luisa Lucie-Smith
Machine Learning the formation of dark matter halos in the Universe
Read More ›


10 November 2020 › Frank Noé
Boltzmann-generating Flows
Read More ›


27 October 2020 › Gábor Csányi
Representation and regression problems for molecular structure and dynamics
Read More ›


13 October 2020 › David Fischer
Attributing variance in single-cell genomics
Read More ›


29 September 2020 › Christoph Weniger
Precision analysis of gravitational strong lensing images with nested likelihood-free inference
Read More ›


15 September 2020 › Erik Henning Thiede
Permutation-Equivariant neural networks for Molecular Generation
Read More ›


30 June 2020 › Tristan Bereau
Physically-motivated machine learning for multiscale molecular simulations
Read More ›