This group has the main goal study and develop new computational approaches with multiples simultaneous evolutionary capabilities (symbolic or sub-symbolic). In fact, engaging Multi Agent Systems (i.e. memetic capabilities like communication and coordination), to evolutionary systems aspects (genetics).
The proposed research includes classical techniques like environment symbolic representation, intentionality, agents actions and objectives, as well as computational intelligence techniques like swarm intelligence, evolutionary computation and cultural algorithms.
We imagine our hybrid approach can effectively solve at the same time problems whose the adaptation capability and execution of semantically qualified actions. Still, we want to make efforts to use a cybernetics approach. With this approach the potential of distributed intelligence with several levels of retro-feed (memetic and genetic) can be used to handle artificial or natural systems.
Theoretic sub-problems rise of this new class of systems. We will investigate some aspects like: (i) communication; (ii) social behaviour; (iii) individual actions semantics; (iv) environment and society structural elements; (v) influence of genetic over memetic; (vi) influence of memetic over genetic; (vii) rise of individual behaviours; (viii) rise of social behaviours.
Pragmatic sub-problems aspects will be the investigated like: (i) environment; (ii) objectives; (iii) agent actions; (iv) society and environment structural elements; and, (v) social behaviors. The expected results will be used to build plausible socio-economic simulations.
Social Computing Team - University of Pernambuco 