ARTIFICIAL COGNITIVE SYSTEM HAVING A PROACTIVE STUDYING FUNCTION USING AN UNCERTAINTY MEASURE BASED ON CLASS PROBABILITY OUTPUT NETWORKS AND PROACTIVE STUDYING METHOD FOR THE SAME

摘要

The present invention provides an artificial cognitive system having a proactive learning function using uncertainty measurement based on class probability output networks, and a proactive learning method for the same. The proactive learning method comprises: a first step for allowing an artificial cognitive control unit to form a body of knowledge for learning materials inputted in a video or audio format through a camera and a microphone and learning materials inputted by wire and wireless through an interface unit; a second step for allowing the artificial cognitive control unit to predict the reliability of a corresponding learning material according to the formed body of knowledge, based on a predetermined proactive knowledge propagation model, and executing proactive learning of a proactive knowledge propagation model type depending on the predicted result, after the first step; and a third step for analyzing proactive learning performance during the first step and supplementing and modifying a calculation model included in the proactive knowledge propagation model depending on the analysis result. The present invention as described above can develop a calculation model having a brain cognitive function by complexly using an information engineering technique and a brain cognitive scientific technique, and can execute proactive learning using uncertainty measurement based on class probability output networks for intelligent robots, thereby allowing the intelligent robots to improve a cognitive function by themselves. [Reference numerals] (12) Video signal processing module; (5) Audio signal processing module; (6) Artificial cognitive control unit; (7) Cognitive learning module; (8) Memory unit; (9) Interface unit