To incorporate the superiority of both stochastic and robust ap-proaches, a data-driven stochastic optimization is employed to solve the security constrained unit commitment model. This ap-proach makes the full use of the historical data information to generate a set of possible probability distributions for wind pow-er output and optimizes the unit commitment under the worst-case probability distribution. However, this model suffers from huge computational burden, as a large number of scenarios are considered. To resolve this problem, a duality-free decomposi-tion method is proposed in this paper. This approach does not require the step of dualization, which can save a large set of dual variables and constraints, and therefore reduces the computa-tional burden. In addition, due to the special structure of the second stage, the proposed method can be decomposed into in-dependent sub-problems that are solved in parallel, which fur-thermore improves the computational efficiency. A numerical study on an IEEE 118-bus system with practical data of a wind power system has demonstrated the effectiveness of the proposal.