Evaluation on Development Capacity of Overseas Chinese Enterprises Based on Improved BP Neural-Networks

摘要

An approach based on improved BP neural-networks is proposed for the evaluation on the development capacity of overseas Chinese enterprise. In order to cope with the slow learning speed and local minimal problems of BP neural-networks, an adaptive learning rate selection method based errors is presented. According to the features of overseas Chinese enterprise, the parameters of input level and number of neurons of hidden level of the BP neural-networks are investigated. Finally, the objective and quantified evaluation result can be achieved from the output neuron.