����Ӱҵ

Upon completing his Master’s degree at ����Ӱҵ College, Eric Langezaal (Machine Learning and Machine Intelligence MPhil 2023) has both made advances in Artificial Intelligence research and fulfilled his lifelong dream of studying at the University of Cambridge.

Keen to come to the Cambridge ever since he was a child, the prospect of spending the bulk of three years away from his home country, the Netherlands, was daunting to 18-year-old Eric. He therefore completed his undergraduate degree in Artificial Intelligence (Kunstmatige Intelligentie) at the University of Amsterdam, where he also progressed to a Master’s programme. However, he intermitted his studies in Amsterdam after attaining a place on Cambridge’s MPhil course in Machine Learning and Machine Intelligence, an opportunity he could not miss.

Eric’s MPhil dissertation, entitled “{On, Off}-the-Grid Data Modelling Using Transformer Neural Processes”, explores one way in which AI can overcome challenges in environmental modelling and make climatic predictions without having to solve a large number of equations individually by “brute force” – a process which requires many computers and is thus highly expensive. Once trained, AI models can make predictions without having to tackle these equations, resulting in much faster modelling.

One challenge to training AI on environmental data is that it comprises both gridded data, where satellites take measurements at regular intervals, and off-the-grid data, measured from facilities such as weather stations which are not evenly distributed. Eric’s thesis focused on developing AI models which can process both types of data simultaneously and efficiently, facilitating the use of this information in large-scale modelling.

“These models are attempting to predict environmental variables, giving us either long-term forecasts over multiple weeks or shorter-term forecasts,” Eric says. “They could eventually be applied to get more accurate environmental forecasting for a fraction of what it would cost now, which would not be possible with traditional methods.”

Eric feels that Caius helped to make his MPhil experience particularly positive. He says: “What made Cambridge so special was my college. I picked Caius because I was really interested in the older colleges and all the culture that comes with them, but also because it’s one of the only colleges that guarantees postgraduate housing on its website. Coming from a city where there’s a housing crisis, a college that said that seemed amazing.

“What I liked by far the most was the dining – being able to go to dinner and see your friends there without having to organise it yourself. I didn’t know about Formal Hall before coming to Cambridge, and I ended up picking the college that does the most Formals and cares about them so much.”

Now back in Amsterdam and completing his Master of Science degree, Eric plans after this year to move into the AI industry, hoping initially to work for start-ups before moving on to larger companies.