Foundations of Explanation for the Design of Computational Systems
– What properties do transparent algorithmic systems satisfy?
– What does it mean to know the behavior of an algorithmic system?
– Which properties make an algorithmic system safe?
My research focuses on formal epistemological and logical foundations of the design of computational systems. My aim is to identify, formulate and check logical properties for algorithmic systems, including those in use in AI platforms. In particular, I am interested in identifying formal characteristics of explainable, transparent and safe AI technologies represented as formal systems. Among the relevant properties of interest for medical applications, there are privacy (of patients’ data) and security (of system’s data, in particular with reference to diagnosis and therapy).
1) G.Primiero, On the Foundations of Computing, OUP, 2020.
2) G.Primiero, A logic of efficient and optimal designs, Journal of Logic and Computation, Volume 29, Issue 6, October 2019, Pages 947–968
3) G.Primiero, F.Raimondi, T.Chen, A theory of change for prioritised resilient and evolvable software systems, Synthese.