Victor Elvira
Edward Isayev has written the following article as part of our series of Academic Interviews; featuring Victor Elvira!
Dr Victor Elvira is a leading researcher in the fields of statistical modelling and inference, and more generally, data science, statistics, and signal processing.
‘What does it mean to be a mathematician?’
While Victor is at the forefront of his field, he didn’t actually start out in maths. He recalls always being interested in the “inner mechanisms of nature”, so when it came time to choosing what to study at university, he hit a crossroads.
At that time, telecommunication was starting to take off, so the natural choice seemed to be engineering. It was the perfect blend of searching for fundamental truths and having an impact on people’s lives, so, as such, he received his PhD in electrical engineering from the University of Cantabria in Spain. It was during this time however, that he found his interests move more towards mathematical sciences, with an interest developing not only in the applications but also in the inner structure of things.
‘Understanding what is hidden to our eyes’
Victor told me about what he personally enjoys in this hunt for “why” and “how”. ‘If you are able to understand the inner mechanisms that are hidden to our eye, then complex phenomena become way simpler than the mess we actually see around us.’ One could take a more pragmatic approach and focus on applications, but, counterintuitively, ‘applications are in that complex world, understanding “why” is more foundational, and can allows us to get back to those applications later and find them much easier.’
‘Then, life becomes more interesting’
Victor explains that completing a postgraduate degree is tantamount to working very hard, generally on a single project, aiming to find, or propose, something new, in one specific area. After you get your PhD, ‘academic life becomes even more interesting, and you can get to investigate a multitude of problems from across a variety of fields’.
His work can broadly be divided into two main areas: Monte Carlo methods for Bayesian statistics, and, what he calls, “complex dynamical systems”. The former being used to help advance the latter.
Monte Carlo methods are, essentially, a class of computational algorithms that are used to estimate unknowns, related to data, in a probabilistic manner. Victor highlights `the importance of quantifying what we do not know in this uncertain world’. These methods are now incredibly popular and widespread due to the surge of Bayesian statistics, with applications also to those complex dynamical systems, being incredibly varied. Some good examples of these systems can be found in climate, energy planning, neuroscience, or ecology, to name a few.
This variety of problems, and their fundamental similarities, are open to those who understand the underlying mechanisms of our world, which are seemingly best expressed in terms of mathematics, and leads to an incredibly dynamic and interesting work environment.
‘Not because of the Weather’
I asked Victor why he chose to come to the University of Edinburgh and, naturally, it wasn’t the weather that compelled him! His main reason was access to good collaborators, ‘you need them, especially if you want to do something impactful’. Between the Bayes Centre, the Schools of Maths, Engineering, and Informatics, the Maxwell Centre, and so on, there is an incredible variety of interesting people to work with in Edinburgh.
‘To be a researcher, you need to be persistent and have unlimited curiosity’
Victor reiterated the importance for collaborators in maths by explaining that ‘you need to accept, even to embrace, that you can learn a lot from colleagues, even more from those in other areas. For example, our design choices when using Monte Carlo methods determine if the computation can be done “in one second, or one week”, and for this, we collaborate with world-wide experts in optimisation’.
Optimisation and statistics are growing closer and closer together and Victor finds himself working within areas that overlap within data science and machine learning, with a strong accent in mathematically sound methods; this is where his interests lie at the moment. ‘By analysing the evolution of mathematics, we now know that its branches are more connected than we thought, and probably way more connected than what we think in 2022.’
‘Do what you love, not what they like’
When asked about advice for aspiring mathematicians, Victor simply stated that one should ‘do what you love, not what they like’, and emphasised how important that is. ‘You need to do something that, when you wake up in the morning, you feel excited and deeply passionate about.’
My final question to Victor was about why he likes maths and he responded, ‘that is an easy one: it allows us to best understand the reality we live in’.