Statistics graduates return for Alumni Showcase

15 October 2016

Landscape image of Associate Professor Renate Meyer in front of a presentation screen.
Associate Professor Renate Meyer discusses the important role statistics played in detecting gravitational waves.

What did statistics have to do with this year’s spectacular discovery of gravitational waves? How does statistics help us eradicate pests like possums, rats and stoats? And how can you wrangle a representative sample out of social networks?

These were just three of the questions posed – and answered – by Department of Statistics academics who presented their research at the first joint Alumni Showcase, a collaboration by the Statistics, Mathematics and Computer Science departments. It took place on Saturday 15 October in the new Science building at the University’s City Campus. The day aimed to strengthen links with graduates by presenting seven short talks on current research, information stands, guided tours of the computer history museum, and a lunch over which graduates had the opportunity to renew relationships with their former lecturers and forge new connections.

Among the visitors was David Grant, who did a masters in pure maths, graduating in 2004, and has since started a postgraduate diploma in Statistics. He's now a data scientist for the Bank of New Zealand. “I really enjoyed the day. It's been really good to hear about some of the latest stuff in stats and maths,” he said. The talk he most enjoyed was on gravitational waves, by Associate Professor Renate Meyer: “I hadn't heard much about the role of statistics in astrophysics.”

And indeed, in her talk, Associate Professor Renate Meyer said that few people realised the “significant role” that statistics played in the “spectacular discovery of gravitational waves” in February. In the 1990s, Associate Professor Meyer and physicist colleague Nelson Christensen developed techniques using Markov Chain Monte Carlo methods – a type of algorithm – to decipher the highly complex signals captured by gravitational wave detectors. These techniques helped researchers extract the tiny and hard-to-detect signals from statistical “noise”.

In his talk, Professor Thomas Lumley pointed out that statisticians had reliable methods for studying a portion of a population and using that to make inferences about an entire population, but were still working on ways to reap useful samples out of social networks. “Maximum likelihood, which is the all-purposes statistical approach, doesn’t work under sampling from networks, and this has been proven,” he said. “So what we are interested in is how to make it work, how to do weighted estimation for sampling frameworks.”

In her talk, Associate Professor Rachel Fewster showed how the CatchIT programme was helping community pest-control groups record and see, with data-driven movie animations, the impact of their efforts to eliminate predators such as stoats, rats, mice and possums. More and more schools were becoming involved, said Associate Professor Fewster, and the programme showed children the value of statistics in everyday life. “We’re trying to build motivation and inspiration with CatchIT – and an understanding of what people are contributing.”

Associate Professor Ilze Ziedins, the head of Statistics, was delighted with the keen interest alumni showed in the talks and presentations. “People were very interested in the inter-disciplinary research in Statistics, Mathematics and Computer Science, and it was great to see so many attendees stay for all the talks.”

Photos from the day are on the Department of Statistics’ Facebook page.