Abstract：Neural Networks (NNs) are used to solving the relational duplicate-removal and sorting problems in this paper. The duplicate-removal and sorting can beefficiently computed in a no learning, shallow, and linear threshold feedforwardNN. These two computations can be finished using only 7 and 6 layers NNs, respectively. They both have the time complexity O(1) which is independent of thecomputation size.
收稿日期: 1994-03-25
引用本文:
高劲松 孟令硅. 基于神经网络的关系消重和排序运算[J]. , 1994, 33(3): 0-0.
高劲松 孟令硅. THE RELATIONAL DUPLICATE-REMOVAL AND SORTING COMPUTATIONS BASED ON NEURAL NETWORKS. , 1994, 33(3): 0-0.
