Meca Discit is an ongoing research project in AGI to develop the analytical model of an unbounded learning system. The motivating context of the project is to add a powerful but lightweight learning capability to the Knowledge Acquisition Phase of autonomous agents whose design is based on the Meca Sapiens Architecture.
The objective is to outline a learning mechanism (termed a Cogitator) that integrates all aspects of learning and discovery into a single coherent structure. The chosen approach is analytical and top-down. Its aim is to use the concepts and symbolic language of Mathematics and Theoretical Computer Science to describe the “essence of learning” in terms of a feasible software architecture.
The research is on going. Your useful comments or contributions are welcome.
Read the project statement HERE.
Read articles about Machine Learning HERE