A Model of Emotion as Patterned Metacontrol

摘要: Abstract Adaptive agents use feedback as a key strategy to cope with uncertainty and change in their environments. The information fed back from the sensorimotor loop into control subsystem can be used four different elements of controller: parameters associated model, model itself, functional organization agent realization agent. There are many alternatives hence complexity agent’s space potential configurations is daunting. only viable alternative for space- time-constrained practical, economical, evolutionary terms achieve reduction dimensionality this configuration space. Emotions play critical role reduction. achieved by functionalization, interface minimization patterning, i.e. selection among predefined set organizational configurations. This analysis lets us state how autonomy emerges integration cognitive, emotional autonomic systems strict terms: closure dependency. Emotion-based morphofunctional able exhibit complex adaptation patterns at reduced cognitive cost. In article we show general emotion supports biological operate following theoretical model. We will also applicability construction wide spectrum artificial systems.