Artificial Intelligence (AI) is revolutionizing the world, from its applications in the advance of science to its use in our daily routines. Therefore, the question about the application of AI to the distributed and secure computing fields is no longer a question of "if", but a question of "when". Our research is trying to bring that "when" to the present.
Follow the news about the team here.
Distributed algorithms are fundamental in modern computation, allowing multiple nodes to work as a single unit of computation, originating what we usually call a distributed system. However, the generation of such algorithms is a laborious, complex, and time-consuming task, where multiple properties must be taken into account, such as the failures to tolerate or the synchrony model of the system. With the recent advances in Artificial Intelligence (AI), one question arises: can AI help in the generation of such algorithms? We are trying to prove that the answer is yes. This research is conducted by the FAULTAGE subteam (FAULt Tolerant Algorithm GEneration).
closed
Check this ResearchGate page to access the publications.
This group is divided into different teams, depending on the purpose. You can see the list of teams here.
To check the current members, see the members list here.
Interested in any topic or with an idea that merges the fields of Distributed Computing, Cybersecurity, and Artificial Intelligence? Send an email to [email protected]