DATAI explores neural networks inspired by nature to understand intelligence

The Institute for data Science and Artificial Intelligence (DATAI) of the of the University of Navarra held a new seminar as part of its scientific lecture series . The session, entitled "Revolution in neural networks: introducing spocu", was given by Prof. Milán Stehlik, researcher at the University of Valparaíso (Chile) and expert in outlier modeling (techniques to analyze rare or infrequent events), artificial intelligence and optimal design of experiments.

Stehlik explained how adaptive transfer functions allow neural networks to automatically adjust the way they process information according to the data they receive. This approach is essential for analyzing complex systems such as the human brain or the behavior of bacteria.

Milán Stehlik during his talk.

From algorithms to the "intelligence of bacterial populations".

The researcher also introduced the idea of "neither human nor artificial intelligence", reference letter to the ability of bacteria or entire ecosystems to make quick and efficient decisions in a matter of milliseconds. This collective behavior, he noted, may inspire new ways to design more adaptive algorithms, applicable in fields such as personalized medicine, biology, engineering or the safety of autonomous systems. "We need more robust tools to respond to the complexity of today's data and learn in dynamic and highly uncertain contexts," said Prof. Stehlik.

The talk included practical examples and encouraged a discussion among attendees on how to translate these ideas into real problems, from risk prediction to the design of personalized medical treatments.

For Jesús López Fidalgo, director of DATAI, "having researchers of international reference letter like Professor Stehlik reinforces our mission statement of bringing data science closer to the great challenges of society, promoting interdisciplinary partnership and frontier research ".