Computation, communication and coordination lie at the core of intelligence, which, in turn, rely on the transmission and transformation of information. Information theory is the scientific endeavor that aims to understand the fundamental mathematical limits of information transmission and transformation and guide the development of algorithms which approach these limits in practical systems.
Hence, we can summarize the fundamental objective of INFORMED-AI as:
- laying the theoretical foundations of learning and intelligence across distributed agents with heterogenous computational capabilities, diverse and private datasets, and limited communication links, and
- developing practical algorithms that will seamlessly scale intelligence across large networks, potentially in the presence of adversaries.
Our 5 year project will contribute to the transformation of this technological landscape and inspire new directions for fundamental AI research using the expertise of our academic and industry researchers, within the structure of our research themes and related work packages.
The research specifically addresses challenges in areas including privacy, security against malicious agents, resilience to network failures, and optimal use of agent heterogeneity.