论文部分内容阅读
针对采购拍卖中存在的品种多、数量大、利润差异大以及评价指标体系不合理等特点,首先在供应商投标后利用自组织神经网络和分包算法对供应商投标进行约减,以提高包内物品相似度、均衡供应商之间的竞争性、降低胜者确定问题算法的复杂度;接着利用数据包络分析中的C~2R模型对“约减”后的投标进行相对评价并确定最终获胜供应商,以解决传统指标体系评价方法中假设属性间不相关以及人为设定权重等不足.该模型为采购拍卖胜者确定问题的解决提出了新的思路,具有很好的实用性.
Aiming at the characteristics of many varieties, large quantities, big difference in profits and unreasonable evaluation index system in the procurement auction, this paper first reduces the supplier bidding by using self-organizing neural network and subcontracting algorithm after the supplier bidding, in order to improve the package The similarity between objects, the competition between suppliers is balanced, and the complexity of the algorithm for determining the winner is reduced. Then the C ~ 2R model in data envelopment analysis is used to evaluate the bidding after “about” To determine the final winner of the supplier in order to solve the traditional index system evaluation method is not related to the assumption between the attributes and the weight of artificial set.The model for the auction auction winner to solve the problem set a new idea has good practicality .