基于粗糙集-BP神经网络的航空风险评价研究

摘要

民用航空是一个高风险的行业，航空系统风险评价具有重大的意义。文章旨在将数据挖掘技术中的粗糙集理论和神经网络相结合，以新的视角去解决航空风险评价问题。该方法主要利用粗糙集软件ROSETTA及MATLAB软件中的神经网络工具箱，实现了粗糙集对数据前置预处理，神经网络模拟评价，并通过实例验证可行。%Civil aviation is a high-risk industry, and the aviation system risk assessment is of great significance. The paper aims to bring in rough set theory and neural network of data mining to solve the problem of the aviation risk assessment with a new perspective. The method mainly using rough sets software ROSETTA and neural network toolbox in MATLAB software, realizes the rough sets fronting data preprocessing and neural networks' simulation evaluation, which is verified feasible through examples.