Abstract: To accurately predict the risk level of coal and gas outburst, we propose a prediction model of gas outburst risk level, i.e., sdPSO-QGNN, based on the quantum gated node neural network (QGNN) and subdimensional evolutionary particle swarm optimization (sdPSO). We employ the gray relational analysis (GRA) to reduce the dimension of the prominent influencing factors and use the selected main control factors as the input of QGNN. Then, we use sdPSO to optimize the parameters of QGNN to improve the global and local search capabilities of the QGNN. We establish an sdPSO-QGNN gas outburst risk level prediction model to predict the gas outburst risk. The experimental results show that, compared with other prediction models, such as BP, SαS-PNN, IPSO-SVM, and Memetic-ELM, the proposed method has a significant effect on the improvement of the generalization capability and prediction accuracy of the model.