Data science

With the new “Data Science” Area, FBK develops the activities of machine learning, complex networks, and deep learning as core competence of the center and application opportunities in big data and predictive analysis.

“Data science” has rapidly become a factor of social and economic transformation: from the big data of organizations to personal sensors, Data Science has the potential to guide the improvement of human and environmental health, providing new tools for knowledge and decision-making. Connecting Data Science and AI opens up opportunities for social and economic change, generating new ideas that can change the world and transforming the ability to decide in a continuous, low cost and high-value process, starting from the combination of complex, multimodal and even incomplete data. Also, the time has come to ensure transparent access and reproducibility to these new technologies, by strongly steering our research actions towards a “Data Science for Good” and offering know-how oriented training opportunities.

In addition to scientific challenges, research in Data Science offers new solutions in fields such as health, quality and production in industry and agriculture, environmental safety, development, social networks and public infrastructures. It is already absolutely essential for sales organizations (e.g. retail), corporate strategies, decision-making processes of public entities. The Data Science Area will therefore contribute new tools to the projects of the ICT Areas dedicated to vertical domains, will increase FBK’s ability to develop projects of ethical importance and local impact, and will leverage its ability to devise analytical solutions and predictive models for organizations and businesses. The topic of deep learning will be put at the center of applied research actions, both for scientific projects and by developing new solutions for clouds, sensors and portable devices. A new action will be dedicated to complex multilayer networks, with a junior group to explore aspects of basic research and applications.

As for technical challenges, we intend to bring the skills of deep learning and geodata science (nowcasting) to the system, develop concepts of predictive analysis associated with dynamic visualization of data, construct the elements connecting data science and AI in genomic research, develop the topic of predictive maintenance in cross-over with the Industry4.0 HII.