企业市场占有率预测神经网络模型

目的研究一种与背景无关，能直接模拟市场选择机制的新的预测市场占有率模型。方法 以神经网络技术为基础来建立模型，采用单隐层神经网络结构，以反向传播算法训练神经网络。结果与结论通过与常用的市场占有率预测模型进行数值计算比较，新模型的预测精度高，可靠性强，使预测效果能得到明显改善。

A Neural Network Model for Market Share Prediction

Aim To present a new market share prediction model that can directly simulate the market choice mechanism without its context Methods A single hidden hierarchical neural network is used to build up a model It uses a back propagation(BP) algorithm to train the neural network Results and Conclusion This new model possesses a much higher precision and reliability when compared with other typical market share models by computed examples