This paper presents an optimization of transonic cantilevered stators based on the combination of sweep and lean in a 5-stage axial compressor to improve its adiabatic efficiency and maintain its total pressure ratio by employing a time-saving integrated optimization design system. The system combines design of experiment with artificial neural network and computational fluid dynamics. The optimization is focused on the first stator stage and the motivation is to understand the mechanism of the performance improvement. It shows that the effect of combinational sweep and lean not only reduces the size of the transonic regions on the blade suction surface near the leading edge but also decreases the leakage loss, thus resulting in lower aerodynamic losses in hub region. Simultaneously, the downstream rotor achieves a higher aerodynamic performance due to a better matching for the improvement of the inlet working conditions. After optimization, the stator loss is reduced by 3.7%, whereas the adiabatic efficiency of the downstream rotor is increased by 0.64%, thus achieving an increment of 0.22% of peak efficiency of the entire compressor. The results show that cantilevered stators with the characteristic of sweep and lean have a significant potential to improve the performance of multistage axial-flow compressor. (C) 2016 Elsevier Ltd. All rights reserved.