Riskcenter members were selected by the jury of the Fundacion BBVA 2019 to develop a research project on “Risk Analytics: turning extremes into core knowledge”.

Risk analysis is about to suffer a huge revolution with the unprecedented access of organizations, governments, citizens and enterprises to amounts and varieties of data coming from a huge array of sources, such as wireless sensors, the internet of things, fast storage and retrieval devices. We will contribute to redefine the field of risk analytics  with algorithms that address a unique combination of challenges in big data, with the presence of extreme events and multivariate dependent dimensions. Our own experience in insurance analytics and statistics helps us identify three areas of interest: (i) modeling the prediction of risk, (ii) preprocessing data to weight outliers rather than to delete information and, finally, (iii) turning classical machine learning models into risk-constrained expectation maximizers, classifiers and optimizers. We think of immediate applications in the prevention of traffic accidents and semiautonomous driving, by creating personalized early warnings when the levels of security measured by telematics are exceeded, but we also aim to foster research with a wider outreach in the context of dynamic identification of patterns that precede the occurrence of catastrophes. Our main task is to recommend how to redesign data collection and modeling processes, by integrating tail instances in predictive tools towards the anticipation of adverse situations. We also aim at redesign the nature of the models that seek to predict risk values rather than average grounded rules. In this way we pursue an added value in the general approach to modern technologies by changing the paradigm that mainly focuses on similarities into the discovery of rubrics within a new approach towards preventing risk. While keeping the speed of algorithms within a sensible range, we propose procedures that will improve the methodology and the technology to create efficient real-time early warning devices that will improve our lives and the safety of organizations.


More information is going to be shared at


Previous winners: