基于ANN—BP模型的电子商务信用风险形成思路

企业生存风险的识别的本质是确定企业生存风险识别的模式和影响企业生存风险状态的各指标权重，但是众多影响因素间不存在确定的函数关系表达式，并且各指标权重的确定也相当复杂。人工神经网络（ANN）基于并行处理机制从结构上对人类的思维过程进行模拟，从而能实现人类思维的某些功能。人工神经网络可以实现任意形式的映射，这就为企业生存风险识别提供了一种新的思路。基于人工神经网络（ANN）的电子商务信用风险模式识别，能够充分利用样本电子商务信用风险的有关信息，通过高度的非线性映射，揭示感知信用风险与其相关影响因素即主要诱因的内在作用机理，从而从根本上克服了感知信用风险测度或识别中建模及其求解的困难。

Thinking on the Formation of E-business Credit Risk Based on ANN-BP Model

The identification of an enterprise＇s survival risk essentially lies in making sure the pattern of survival risk identification and the weights of various indexes impacting the state of the enterprise＇s survival risk. However, no determinate functional relation has been found among numerous impacting factors, and it is quite complicated to ascertain the weights of various indexes. A~ficial neural network （ANN）, based on parallel processing mechanism, simulates the thinking process of mankind from the structural perspective, and thus achieves certain functions of human thinking. ANN can carry out mappings in any forms, and hence provides a new way of thinking towards an enterprise＇s survival risk identification. ANN-based identification of E-business credit risk pattern is able to get the most out of the information related to samples＇ E-business credit risk, and reveal the inherent mechanism of the effects of relevant impacting factors or major inducements on perceived credit risk, through high-level non-linear mapping, and thus fundamentally overcomes the digicuhies existing in modeling for perceived credit risk measurement and identification, as well as its solution.