engleski

An Approach to Universal Interaction on the Case of Knowledge Transfer

This paper presents an approach to universal interaction which can be used for robot knowledge acquisition and transfer of acquired knowledge between different robots. There are similarities between human and machine learning techniques so learning by demonstration and conceptual learning were used as a basis for demonstrating our proposed type of interaction. Advantages and limitations of proposed interaction are described and discussed. Also, empirical study to test our approach was carried out and results are presented and analyzed.

learning ; artificial intelligence ; machine learning ; universal interaction ; knowledge transfer

nije evidentirano