The rise of machine learning (ML) has led to an explosion in potential strategies which may be used to learn from data in order to make scientific predictions. For physical scientists who wish to apply ML strategies to a particular domain, the vast number of strategies available has made it difficult to make an a priori assessment of what strategy to adopt. This is further complicated when similar domains have not been previously explored in the literature.
Recently, CHAMPS PDRA Dr. Lars Andersen Bratholm and collaborators worked with Kaggle to design a competition which encouraged data scientists around the world to develop ML models for predicting pairwise nuclear magnetic resonance (NMR) properties for synthetically relevant chemical compounds.
The success this strategy has cultivated highlights the potential of crowd-sourced ML approaches across a range of scientific domains and the CHAMPS symposium “Crowd-sourcing machine learning in NMR” took place on the 5th of March to communicate this message to researchers in the Bristol area.
The symposium started off with Professor Craig butts and PhD student Will Gerrard from the University of Bristol presenting how modelling of NMR properties with DFT and ML can accelerate the drug design process. This was followed by Addison Howard from Kaggle introducing the Kaggle platform, background to the company and interesting findings from the CHAMPS competition. The early non-technical session ended with Lars Bratholm conveying the main findings of the competition from the organizers point of view, namely what can be gained from combining all the different approaches, and how the collaborative environment of Kaggle is something that we can learn from in academia.
The afternoon session turned more technical and featured representatives from six of the top performing teams who all presented their approaches to the competition. The talks of Brandon Anderson, Luka Stojanovic, Milos Popovic, Sunghwan Choi, Andres Torrubia and Devin Wilmott all spurred great discussion with audience, which continued the following day with the competition organizers.
Dr. Lars Andersen Bratholm.