Data Envelopment Analysis (DEA) is one of the best methods for measuring the efficiency and productivity of Decision Making Units (DMU). Evaluating the efficiency of DMUs which have two or several stages by using the conventional DEA models, is equal to consider them as black box. This method, omits the effect of intermediate measure on efficiency. Therefore, just the first network inputs and the last network outputs will be effected on network efficiency. After solving each DEA model, in order to improving the DMUs efficiency, introduce one efficient market to evaluated DMU is so important. Conventional DEA network models cant ensures that the market of each evaluated is efficient .In this paper, by considering DMUs that have two stages network process, in order to find projection frontier point some models have been proposed. These models after finding the best intermediate product corresponded to each network, found the overall network efficiency and introduce one projection point for each DMU. تفاصيل المقالة