This paper describes a multidisciplinary design optimization for performance improvement of an electric-ducted fan rotor using free-form deformation (FFD) and data mining techniques. A practical partitioning approach for FFD parameterization was applied in combination with engineering design parameters to optimize the fan rotor. Regression analysis was used to initially determine an approximation function for the blade static stress and subsequently integrated into a fully coupled iterative loop to optimize the blade considering two operating points. Two optimization solutions for 10 and 12 blades were performed. Percentage improvements in the efficiency of 1.05% and 1.32% were realized for 10 and 12 blades, respectively, at near peak efficiency flowrate. Also, blade static stress was reduced by percentages of 5.49% and 12.37% for 10 and 12 blades compared with the baseline. Data mining results revealed key design variable sensitivities where blade twist, sweep, chord, and hub thickness distribution were found to be the most influential for 12 blades while for 10 blades, blade lean, sweep and chord at the midspan and tip. The optimized blades were found to have a significant increase in chord from midspan to tip mimicking a wide chord fan blade particularly for 10 blades. Analysis of the flow field revealed that the axial velocity from 0.4 to 0.8 spanwise length increased significantly for the optimum blades due to the increase in blade twist and chord length at all stable operating points. However, the leakage trajectory relative to the blade chord was observed to be larger and interacted with the trailing edge wake flow downstream for the optimum blades at near-stall conditions. Furthermore, the increase in chord length and the thinning of the blade close to the trailing edge from 0.4 to 0.8 span reduced the suction-side blade loading and static stress.