THIS WEEK, IN THE ‘BEHIND THE SCENES’ SERIES, WE’RE SPEAKING WITH PROFESSOR OF MACHINE LEARNING, AT THE BASQUE CENTRE FOR APPLIED MATHEMATICS (BCAM), NOVI QUADRIANTO. NOVI WILL BE TELLING US ABOUT HIS ROLE AND THE ROLE OF BCAM WITHIN THE TANGO PROJECT.

Thank you for speaking to us today. Could you start by introducing yourself and telling us where you are based?

I am a Professor of Machine Learning at the University of Sussex and a Senior Researcher at BCAM. My involvement with TANGO is through our Severo Ochoa Strategic Lab on Trustworthy Machine Learning. The Strategic Lab is a joint research lab between BCAM (Spain), and the University of Sussex (UK) established in 2021. This Strategic Lab programme highlights the importance of international collaboration between BCAM and universities. Its goals are, to promote collaborative research on topics such as machine learning methods, and thus strengthen BCAM’s research areas and generate synergies, to explore the connection between pure/applied mathematics and new research topics in collaboration with BCAM research areas, to strengthen contacts within the international scientific community and to provide specialized training to the members of the Lab.

How would you describe TANGO with three words only?

Human-AI Collaborative Intelligence. That is 3 words, thanks to the hyphen!

What does your typical working day in the TANGO project involve?

It is never ordinary, our Lab is in Spain, the team members are currently in Spain, UK, and Germany so we are having a mini consortium within the TANGO consortium. We have a lot of brainstorming sessions via Zoom and we hope to leverage our Lab setup so that our remote team members can non-remotely facilitate our collaborations with other TANGO partners from 9 different European countries.

 What is your main task in the TANGO project?

My main tasks are leading the development of reliable models with fairness and bias considerations, translating those reliable models into ethical and trustworthy decision-making, and contributing to case studies.

What do you like most about your role?

What I like most about my role in the TANGO project is the opportunity to contribute to creating reliable models that prioritize fairness and consider biases. It’s fulfilling to know that my work is helping to ensure that decision-making processes are ethical and trustworthy. Additionally, being able to contribute to real-world case studies allows me to see the tangible impact of our efforts in improving systems and outcomes.

How do your professional interests match the objectives of TANGO?

My research interest and expertise lie in machine learning, with an emphasis in ethical and trustworthy machine learning (auditing/mitigating inappropriate bias against protected subgroups and improving transparency of algorithmic systems), safe and robust machine learning (ensuring reliably good performance even when encountering extreme situations), and interactive machine learning (facilitating an understanding between a user and an algorithmic system). They match with the objectives of TANGO.

What is unique about TANGO in your opinion?

To achieve truly Human-AI collaborative intelligence, we need collaborations between cognitive science experts (who study human mind and brain) and computer science experts (who study AI), and this is exactly what TANGO delivers. Back in 2017, I promised UK grant reviewers a follow-up project of EP/P03442X/1 that would require interdisciplinary research teams comprising machine learning experts and cognitive scientists. I feel privileged to be able to deliver on the promise and be part of this excellent TANGO consortium.

What makes your organisation ideal for participating in the research/activities of TANGO?

BCAM is an ideal organization for participating in the research and activities of TANGO due to our expertise in advanced mathematics and data science. In addition to the EU projects in the topics of machine learning and algorithmic fairness in which BCAM participates (BayesianGDPR and Act.AI for instance), the center has been awarded (2013, 2018, 2022) with the Severo Ochoa distinction by the Spanish Government, the highest possible recognition of a research center in Spain. BCAM’s goal is to lead the development of Mathematics, contributing to scientific and technological development oriented towards social welfare. The interdisciplinary groups of BCAM are focused on understanding and improving AI/Machine Learning algorithms employing our expertise in advanced methods from Data Analysis, AI, Applied Mathematics, Fourier and Functional Analysis, and Partial Differential Equations (PDEs) among other research areas.

Has anything surprised you in the first months of the project?

Very good atmosphere and efficient support on behalf of the coordinators of the project.

What do you see as the biggest challenge for TANGO?

Achieving a virtuous cycle of cognitive-computational research, in which computational research results will be translated to working software systems and large case studies, which in turn will provide feedback for cognitive theory refinement on human-AI collaborative intelligence.

What does TANGO have in common with other EU-funded research projects?

Firstly, like many EU-funded projects, TANGO aims to address societal challenges by leveraging scientific research and technological advancements. It aligns with the European Union’s priorities for fostering innovation, sustainability, and social inclusion. Additionally, TANGO, operates on a collaborative basis, bringing together researchers, institutions, and stakeholders from across Europe and beyond. This collaborative approach encourages knowledge exchange, fosters interdisciplinary cooperation, and enhances the collective expertise needed to tackle complex problems effectively. Furthermore, TANGO, like other EU-funded projects, is committed to promote ethical considerations, transparency, and strict accountability in research and development activities. It strives to adhere to principles of responsible innovation, ensuring that its outcomes benefit society while minimizing potential risks and negative impacts.

What is one key thing you have learned from working on TANGO so far?

An excellent coordinator is crucial for a large multi-disciplinary project, Andrea Passerini and the research team at UNITN have done amazing jobs with regular brainstorming sessions.

Could you describe the overall expected impact of the TANGO project in three words?

Collaborative Intelligence RIA. This time, thanks to the abbreviation of RIA as Research, Innovation and Action!

What would be your advice to anyone interested in getting involved with a Horizon Europe project?

Keep yourself updated on the latest calls for proposals, funding opportunities, and thematic priorities under Horizon Europe. In addition, collaboration is key in Horizon Europe projects. Seek opportunities to network with potential partners from academia, industry, and other sectors who share your research interests and expertise and try to build strong partnerships because they will enhance your project proposal and increase its chances of success.

Lastly, invest time and effort in developing a high-quality project proposal and seek support and guidance from your institution’s research services and colleagues. And of course, be persistent apply lessons learnt from previous experiences and improve your proposals to increase your chances of success.

Thank you for taking the time to speak to us, Novi. We look forward to more updates from the team at BCAM. 

To keep up to date with the Behind the Scenes series and all TANGO updates, make sure to follow us on TANGO (@TANGO_horizon) / X (twitter.com) and TANGO LinkedIn