Skip to main content


This area focuses on the study of systems, algorithms and models that enable computers to learn and take decisions based on data without requiring any explicit programming.

A clear example of this approach is deep learning which is currently the basis of a number of industrial and commercial applications in the Artificial Intelligence field.

In the medium to long term, research in this area aims not just to improve existing algorithms in this area, but also to create new algorithms that take into account disruptive developments, such as quantum computing, which represents a paradigm shift, and the need to address particularly complex challenges in fields such as biotechnology and personalised medicine.

Explainable AI

The explainability of an AI system is vital to increase transparency and reliability in this system.

Research in this field will serve to understand how decisions AI systems make decisions and provide understandable explanations to users in critical applications such as medical care.


Computational Neuroscience uses mathematical and computational models to understand the development and functioning of the nervous system.

This field of work is decisive for understanding brain principles and tackling complex problems in AI, such as computer vision and natural language processing.

AI in Dynamic

AI can be used to model dynamic systems in various ways, from solving differential equations to using automatic learning to represent complex systems.

This modelling ability is essential in the research, design and control of physical systems in a wide range of applications, where AI-based controllers adjust the system’s inputs to achieve a desired behaviour. This approach proves useful in applications that range from the control of industrial processes through to autonomous vehicles and robotics.


Reinforcement learning is crucial for machines to take autonomous decisions and learn by interacting with their environment.

It can be applied in robotics, games, route optimisation and other areas in which the taking of sequential decisions and adaptation to changing situations are required.

Multimodal AI

Multimodal AI focuses on understanding and combining data from multiple sources, such as text, image and voice, for a fuller and richer understanding.

This approach can improve the understanding of information in applications that involve data of multiple modalities, such as fraud detection, automatic translation and customer service.