An efficient combination strategy for hybrid quantum ensemble classifier

Quantum machine learning has shown advantages in many ways compared to classical learning. In learning, a difficult problem is how learn model with high robustness and strong generalization ability from limited feature space. Combining multiple models as base learners, ensemble (EL) can effectively improve the accuracy, of final model. The key EL lies two aspects, performance learners choice combination strategy. Recently, quantum (QEL) been studied. However, existing strategies QEL are inadequate considering accuracy variance among learners. This paper presents hybrid framework that combines advantages. More importantly, we propose an efficient strategy for improving classification framework. We verify feasibility efficiency our by using MNIST dataset. Simulation results show not only higher lower than single without ensemble, but also better majority voting weighted most cases.