Cognitive foundations of knowledge science and deep knowledge learning by cognitive robots

Recent basic studies reveal that novel solutions to fundamental AI problems are deeply rooted in both the understanding of the natural intelligence and the maturity of suitable mathematical means for rigorously modeling the brain in machine understandable forms. Learning is a cognitive process of knowledge and behavior acquisition. Learning can be classified into five categories known as object identification, cluster classification, functional regression, behavior generation, and knowledge acquisition. The latest discovery in knowledge science by Wang revealed that the basic unit of knowledge is a binary relation (bir) as that of bit for information and data. A fundamental challenge to knowledge learning different from those of deep and recurring neural network technologies has led to the emergence of the field of cognitive machine learning on the basis of recent breakthroughs in denotational mathematics and mathematical engineering. This keynote lecture presents latest advances in formal brain studies and cognitive systems for deep reasoning and deep learning. It is recognized that key technologies enabling cognitive robots mimicking the brain rely not only on deep learning, but also on deep reasoning and thinking towards machinable thoughts and cognitive knowledge bases built by cognitive systems. Fundamental theories and novel technologies for implementing deep thinking robots are demonstrated based on concept algebra, semantics algebra and inference algebra.