Quality-Complexity-Task Universal Intelligence Measurement

The phenomenon of crowd intelligence widely exists in human society.As human beings enter the network age, the phenomenon of crowd intelligence is becoming more extensive and complex. These intelligent agents (i.e., people, enterprises, government, intelligent equipment and goods)connect each other and form a large number of crowd network system.With the development of machine learning techniques, the crowd network is becoming more and more intelligent and autonomous in the physical space. Due to complexity and heterogeneity of the agent in the crowd network, how to reasonable optimize and evaluate the intelligence of crowd network is becoming a very important problem. In order to solve this problem,we propose a formalized and accurate intelligence measurement method named Quality-Complexity-Task(QCT) universal intelligence measurement. Like the human IQ test,we use a kind of intelligence test method to measure agent intelligence. To model this process of test, we design QCT agent-enviroment framework in which the agent and enviroment could interact with each other. We measure the intelligence of agent by evaluating the agents accumulated performance during the intelligence test. Our experiment demonstrate that the method of QCT can achieve the measurement of agent in the crowd network.