商业研究
商業研究
상업연구
Commercial Research
2008年
3期
64~68
,共null页
网络营销绩效 指标体系 神经网络 BP算法
網絡營銷績效 指標體繫 神經網絡 BP算法
망락영소적효 지표체계 신경망락 BP산법
network marketing performance; index system; neural network; BP algorithm
在研究网络营销本质及其绩效特征的基础上,建立企业网络营销绩效评价指标体系并采用BP神经网络建立评价模型。为了克服传统BP神经网络容易陷入局部极小、收敛速度慢的缺点,通过采用LM算法和学习速率的自适应调整来建立新型的BP神经网络,利用模糊综合评价法生成训练样本。该评价模型有效地利用了BP神经网络的优点,避免了评价过程的人为失误。仿真试验表明,该模型取得了较好的效果。
在研究網絡營銷本質及其績效特徵的基礎上,建立企業網絡營銷績效評價指標體繫併採用BP神經網絡建立評價模型。為瞭剋服傳統BP神經網絡容易陷入跼部極小、收斂速度慢的缺點,通過採用LM算法和學習速率的自適應調整來建立新型的BP神經網絡,利用模糊綜閤評價法生成訓練樣本。該評價模型有效地利用瞭BP神經網絡的優點,避免瞭評價過程的人為失誤。倣真試驗錶明,該模型取得瞭較好的效果。
재연구망락영소본질급기적효특정적기출상,건립기업망락영소적효평개지표체계병채용BP신경망락건립평개모형。위료극복전통BP신경망락용역함입국부겁소、수렴속도만적결점,통과채용LM산법화학습속솔적자괄응조정래건립신형적BP신경망락,이용모호종합평개법생성훈련양본。해평개모형유효지이용료BP신경망락적우점,피면료평개과정적인위실오。방진시험표명,해모형취득료교호적효과。
On the basis of research into the essence and performance of network marketing, this paper aims at establishing the index system of evaluation on the performance of network marketing and establish evaluation model through BP neural network. Aiming at overcoming the slow convergence rate and its encountering local minimum of traditional BP neural network. It establishes a new BP neural network. The method of improving traditional BP neural network has been stated by using LM algorithm, adaptive learning rate and produce training swatch of network through a fuzzy and comprehensive evaluation method. The evaluation system utilizes the excellences of BP neutral network effectively and avoid some subjective mistakes in the course of evaluation. The result from the simulation shows the satisfying effect.