The gas pipeline network is a critical infrastructure connecting downstream customers and upstream sources, and the gas supply condition of the gas network directly impacts people's lives. In this paper, a novel methodology is proposed to evaluate the short-period supply reliability of a gas pipeline system based on transient operation optimization and support vector regression. Firstly, supply and demand uncertainty characteristics are analyzed, and representative scenarios are selected through the improved Latin Hypercube Sampling method. Secondly, the transient peak-shaving characteristics are studied, and the physical transient peak-shaving operation optimization model is established to evaluate the satisfaction rate of gas customers and the system under representative scenarios. Then the Support Vector Regression model based on the improved Particle Swarm Optimization is established to predict the satisfaction rate of the customer or system under each random scenario, and the probabilities under different satisfaction degrees of the customer and system are given to reflect the supply reliability of a gas network. The proposed integrated methodology is verified by a specific gas pipeline system, and the results show that the maximum and the mean absolute error of the predicted satisfaction rate of the system's demand quantity can be reduced to 0.0070 and 0.0009, and the time consumed for the evaluation process can be reduced by 95 % compared to the traditional pure physical model. The proposed methodology can lay a solid foundation for the grasp of the short-period supply reliability of gas networks.