Hybrid learning incorporating neural and symbolic processes

Abstract

To develop autonomous agents for sequential decision tasks in a highly reactive fashion, we present a learning model CLARION, which is a hybrid connectionist model based on the two-level approach proposed in the CONSYDERR architecture. The model learns and utilises procedural and declarative knowledge, tapping into the synergy of the two types of processes (subconceptual and conceptual). It unifies neural, reinforcement, and symbolic methods to perform online, bottom-up learning. Experiments in various situations are reported that shed light on the working of the model and demonstrate the performance advantages of the model.