Statistical Science


Dr Alex Beskos awarded a Philip Leverhulme Prize

11 November 2014


Dr Alex Beskos has been selected to receive one of this year's Philip Leverhulme Prizes. These awards recognise the accomplishments of researchers who have already achieved international success in the early stages of their careers. Prizewinners get £100,000 to spend on their research.

Dr Beskos said:

Many statistical models used in diverse scientific fields involve high-dimensional parameters, learning of which guides decision-making. Such problems are naturally cast in a Bayesian framework whereby resolving them corresponds to probing posterior distributions, combining mathematical models with observations. 

My recent investigations can be framed within this field  (related with important applications in inverse problems, filtering and econometrics) whereby standard computational methods are considered ineffective but recent developments have shown great promise for carefully designed algorithms. I have used methodological advances in computational statistics and produced new algorithms to achieve accurate principled (likelihood-based) parameter estimation in the designated directions.

The prize will allow me to extend collaboration and interaction with experts in the field of computational statistics and filtering in the UK and abroad, appoint a post-doctoral researcher, greatly facilitating the development of the required methodology, and participate in important international conferences to keep up with the developments in the field.

I am extremely pleased (of course) I was awarded the prize, as it provides me with a reassurance that my contributions are appreciated in the field.

The prize will allow me to further develop research ideas and contribute in the area of computational statistics.