Selective Ensemble Modeling Method Based on Random Vector Functional Link Network and Game Theory

摘要

Ensemble learning can deal well with the concept drift problem in industrial process data. However, the accuracy and diversity of base learners have a great impact on the generalization performance of the ensemble model. In order to further improve the ability of ensemble model to cope with concept drift, this paper proposes a selective ensemble modeling method based on Random Vector Functional Link network (RVFL) and game theory. RVFL network is used as base learner, the accuracy of the base learner and the contribution rate of the base learner to the diversity of the ensemble model are regarded as the two sides of the game. Game theory is used to solve the optimal selection scheme for the accuracy and diversity of the ensemble model. The rationality and effectiveness of the proposed algorithm are verified by using open data sets and real industrial data.