Measuring the machine intelligence quotient (MIQ) of human-machine cooperative systems

Abstract

We present a practical and systematic strategy for measuring machine intelligence quotient (MIQ) of human-machine cooperative systems. So far, much research related to intelligent control has been established, but the issues about definition and measurement of machine intelligence are not yet clearly formulated. We propose the intelligence task graph (ITG) as a modeling and analysis tool. By using the ITG, the machine contribution of human-machine cooperative systems can be explicitly separated from the human contribution and directly formulated by equations. Thus, the ITG is a very useful modeling tool for estimating the MIQ that was introduced as the index of machine intelligence. We propose a new human-oriented method for estimating the MIQ. This research will help engineers assess intelligent machines that are designed to perform human-like qualities with a high degree of autonomy.