ICML is considered as one of the top AI conferences in the world, and the outstanding paper is awarded to papers of the highest quality among accepted papers and is highly selective. Among 12176 submitted papers this year, only six were selected for this prestigious award.
The paper, titled “Score Matching with Missing Data”, was co-authored by the PhD student Josh Givens (based in the EPSRC-funded Compass CDT), Associate Professor Song Liu (School of Mathematics, University of Bristol), and Dr Henry W. J. Reeve (Nanjing University). Their research introduces two innovative methods of score matching—a key machine learning technique—to a setting where data are partially missing.
Score matching learns a “score model” from data and underpins many modern AI applications, most notably, generative models.
However, the traditional score matching requires fully observed data, which is often not realistic in applications when samples are corrupted. The Bristol team overcomes this issue by proposing “marginal score matching” where the method actively “impute” data from observed samples when training the score model. The research opens a wide range of opportunities for real-world applications where the samples are partially missing, such as images collected from a noisy environment or genomics samples with degraded DNAs/RNAs.
“This award is a fantastic recognition of Bristol’s research strength on cutting-edge, foundational AI research in a competitive global stage,” said Dr. Song Liu. “Our research enables machine learning methods to work reliably in the imperfect data settings we often face in the real world.”