上市公司综合经济效益比较的神经网络方法

A Neural-Network-Based Approach to Comparing Synthetic Economic Benefits of Stock Companies

上市公司综合经济效益排序是个多指标比较问题.作者提出一种基于人工神经网络的方法,通过学习、推理确定权重,较好地揭示样本各指标间的关系,从而对样本进行综合比较和排序;并通过网络的推理求出各指标对总权重所作出的贡献.

Ordering the synthetic economic benefits of stock companies consists of a system of comparing multiple indices. This paper suggests a method based on artificial neural network. The weights of indices can be obtained and the relationships of indexes can be revealed by identification and inference of the relevant knowledge, thus the order of stock companies can be determined. The devotes to weights by indexes can be shown by inference.