Congratulations to STOR-i student Kes Ward, and collaborators, whose paper was selected for discussion as the prestigious JASA Applications & Case Studies invited paper session at JSM 2024.

Each year, the editors of the Journal of the American Statistical Association (JASA) select two papers for presentation at the Joint Statistical Meetings (JSM), the largest gathering of statisticians held in North America: one for the Theory and Methods Invited Session, the other the Applications and Case Studies Invited Session.

In explaining the paper, focussed on finding gamma-ray bursts (flashes of light from exploding stars) using detectors on a satellite in space, Kes said: “Because our data is updating a thousand times a second, and we don't know how long a gamma-ray burst will be before we find it, it can take too much computing power to process all of our signal intervals for statistical checking. People usually get around this by approximating. Using recursive statistical methods, we were able to massively lower the amount of computing power needed to check the signal exactly, which is exciting!”

The paper was written in collaboration with Kes' PhD advisors (Idris Eckley and Paul Fearnhead), together with astrophysicist Giuseppe Dilillo who is based at the University of Udine in Italy.

Following the presentation of the paper at JSM 2024, Kes spoke about the opportunity: “The American Statistical Association chose this paper for discussion because of our close collaboration with astrophysicists driving immediate impact from the research, and our focus on finding statistical methods that scale well to large, fast-moving datasets, which has wide applicability to lots of other real-world problems.”

If you are interested in learning more, read the paper here.