Statistical ‘rockstar’ Brad Efron opens Leiden statistics centre
He is said to be one of the greatest living statisticians. Bradley Efron gave a lecture in Marekerk and opened LUXs, the Leiden University Center for Statistical Science, the first university statistics centre in the Netherlands.
It is the third day of the statistics conference to mark the 60th birthday of statistician Aad van der Vaart. The turnout in Marekerk is high and diverse. Next to me is a Chinese postdoc from the University of Göttingen and in front of me sits Els Goetghebeur from Ghent. She has experienced today’s star close up in Stanford. She knows that Bradley Efron (1938) was one of the first to read all the Harry Potter books, but, more importantly, that he always communicates clearly and does all of his own research. ‘If he has an idea, he programs and analyses datasets himself. Others get students to do that. With his hands-on mentality, he’s not just strong in the theory but is equally at home in the practice.’
No institute yet
Jacqueline Meulman, Professor of Applied Statistics at Leiden University and Stanford University, introduces the soon-to-be-opened LUXs. ‘The first master’s programme in statistics in the Netherlands began in Leiden in 2009. It is a collaboration between different faculties in Leiden, and other universities also participate. Now about 45 students start the programme each year. But there is no institute for statistics in the Netherlands. We are addressing this shortcoming with LUXs.’
As we arrive, Saskia le Cessie, Professor of Medical Statistics at the LUMC, explains why such an institute is important. ‘Statistics researchers are spread throughout the University, but taken together we have the largest statistics group in the Netherlands. This group will now become more visible, and collaboration will become more of a given.’
Extremely useful for big data
Le Cessie is also enthusiastic about Efron. His greatest achievement is the bootstrap method, which is very useful for analysing big data. ‘Genetic data, for instance. Imagine you have measured lots of gene expressions. You can then use bootstrapping to predict whether someone will contract cancer, or if the cancer will spread if they already have it.’
Before Efron takes the floor, Meulman shows us a photo of him looking sour-faced as he stands next to former president, George W. Bush. ‘He was pleased with the prize that he had been awarded, the National Medal of Science, but not with the president.’
When it is his turn at the lectern, Efron says, ‘In retrospect, he wasn’t such a bad president after all – if you look at our current president.’
Ribbon rather than a flashy film
Efron then uses his famous bootstrap method to explain the essence of statistical estimation and prediction. Even for those who are unfamiliar with the subject, the examples hit home. Infant mortality in an African hospital, the effect of cholesterol medicine and the predictability of prostate cancer on the basis of a few thousand (!) gene expressions.
This makes the relevance of statistics all the more clear. And it’s no great tragedy that there is no flashy film or other grand show: after the lecture, Meulman takes one end of a blue ribbon, Aad van der Vaart takes the other, and Efron cuts it with a pair of scissors and congratulates everyone on the new LUXs.
Text: Rianne Lindhout
Photos: Marc de Haan
Mail the editors