摘要
为了提高尘肺病的预测准确性,针对尘肺病历史数据少、不确定的特点,采用多种数据挖掘技术进行建模,提出一种基于GM-BPNN的尘肺病组合预测模型。首先利用灰色模型GM(1,1)对尘肺病进行预测,然后采用BP神经网络对GM(1,1)预测结果进行修正,并采用遗传算法优化BP神经网络的初始权值和阈值,最后对1981~2006年的尘肺病例进行仿真测试。仿真结果表明GM-BPNN很好地解决了尘肺病预测过程中的小样本、非线性问题,相对于单一预测模型,提高了尘肺病的预测精度。
Aiming at the problem of fewer historical data with uncertainty characteristics,in order to improve the accuracy of prediction for dust-pulmonary disease,the paper proposes one prediction model based on the GM-BPNN by using many data mining technology.Firstly,GM(1 ,1 )is used to predict the dust-pulmonary disease,and then BP neural network is used to modify the prediction results of GM (1 ,1 )which initial weights and thresholds of the BP neural network are optimized by genetic algorithm,and finally the test for dust-pulmonary disease case from 1981 to 2006 is conducted.The simulation results show that GM-BPNN is a good solution to the problems of small sample and nonlinear and the proposed model improves the precision of prediction for dust-pulmonary disease.
出处
《微处理机》
2014年第3期52-55,共4页
Microprocessors
基金
河南省科技计划重点项目(102102210416)
关键词
尘肺病
灰色模型
BP神经网络
遗传算法
仿真
Dust-pulmonary disease
Grey model
BP neural network
Genetic algorithm
Simulation
