基于粗集和神经网络的煤炭资源资产分类研究

The Study on the Classification of Coal Resources Assets Based on Rough Sets and ANN

准确的煤炭资源资产分类是有效进行煤炭资源资产管理的前提和基础.针对现有分类方法存在的不足,提出了基于人工神经网络(ANN)与粗集理论(RS)的煤炭资源资产分类方法.首先,由资产分类的样本数据形成决策表,使用专家离散法对数据进行离散处理;然后,采用遗传算法(GA)对决策表进行属性约简;最后根据约简后的属性集构建起煤炭资源资产分类的神经网络模型.实例运行表明,所提出的模型方法比单纯的ANN方法在学习效率和分类准确率方面均有所提高.

The evaluation of coal resources assets is the important and basic work to carry out asset style management of coal resources, and the reliability of classification of coal resource assets is the precondition to complete the work effectively. In this paper, Artificial Neural Networks and Rough Sets theory were integrated to solve the classification problem. The decision table is formed according to the data of the samples, Experts discrete method is used to＇discrete the values of the attributes, GA arithmetic is used to reduce the attributes, and Artificial Neural Networks model is built according to the reduced attribute set. The operation of the example indicated that the proposed model is improved in studying efficiency and classification accuracy than the simple ANN method.