To improve overall performances of transonic compressor rotor, the multi-objective optimization design of the blade profile was carried out by using the combination method of artificial neural network and genetic algorithm based on full three-dimensional optimization design platform. The optimization target is to increase the pressure ratio and efficiency by keeping the rate of flow invariant fundamentally. Results show that corresponding to the design point, the efficiency and pressure ratio of opt1 increase 0.81 % and 1.55 % respectively compared with those of the original blade profile, and the efficiency and pressure ratio of opt2 increase 0.36 % and 3.09 % respectively, and the stall margin of opt1 and opt2 increase 1.58 % and 0.89 % respectively. Therefore, the blade load distribution can be regulated and controlled by decreasing at 50% blade height and increasing the stagger angle at 95 % blade height, and adjusting the wedge angles of leading edge or trailing edge in suction surface, in turn, which improves overall performances of rotor.