高阶前馈神经网络在区域环境质量评价中的应用研究  

Application of higher order feedforward neural networks in regional environmental quality assessment

作　　者：王阿明[1] 刘天放[1] 王绪[2] 

机构地区：[1]中国矿业大学资源与地球科学学院,江苏徐州221008 [2]徐州医学院附属医院放射科,江苏徐州221002

出　　处：《徐州医学院学报》2003年第2期110-113,共4页Acta Academiae Medicinae Xuzhou

摘　　要：目的　探索高阶前馈神经网络模型特性 ,将其应用于区域环境质量评价。方法　通过在多层前馈神经网络中增加高阶连接权建立高阶前馈神经网络模型 ,用此模型研究区域环境质量评价 ,并与传统BP网络的应用结果进行对比。结果　高阶前馈神经网络模型应用于区域环境质量评价时 ,其性能指标优于传统BP网络。Objective To analyzehe characteristics of higher order feedforward neural networks model and demonstrate the application of the model in regional environmental quality assessment. Methods A new way to investigate regional environmental quality assessment by using higher order weights in multilayer feedforward neural networks was provided. The regional environmental quality assessed by using higher order feedforward neural networks model was compared with that by traditional BP model. Results The properties of higher order feedforward neural networks model were superior to those of the traditional BP model when applied in the assessment of regional environmental quality. Conclusion The fine features enable the higher order feedforward neural networks model to be of good prospects in regional environmental quality assessment.

关 键 词：高阶前馈神经网络 模型 区域环境质量 

分 类 号：Q61[生物学—生物物理学]