中文摘要：

为解决现有风险投资项目组合选择方法未考虑资本不可细分和投资收益率难以预测的问题,首先在借鉴MV模型思想的基础上结合风险投资项目评价指标体系定义了项目组合偏好率及其方差,并将二者分别视为项目组合的产出与投入,然后基于DEA相对效率评价原理构建了能够评价出具有Pareto最优性的项目组合优选模型即VCP-DEA/AR模型,最后通过一个数值验证案例证明了该模型的科学有效性和应用可行性。

英文摘要：

To solve the problems of current VC （venture capital） project portfolio selection methods, resulting from the capital indivisibility and the scrutability of VC projects＇ ROI （returns of investment）, the ratio of project portfolio preference（TPR） and its variance （VTPR） were defined based on the project evaluation criteria and the technical thoughts in mean-variance model. After that, the TPR and the VTPR were regarded as outputs and inputs of portfolios respectively, and a model named VCP-DEA/AR was established by the principle of data envelopment analysis. The presented model is able to derive pareto-optimal project portfolios. Applied to an illustrative example, the presented model is proved to be scientific, efficient, and well applicable finally.