一种改进的基于加权模式的BP算法

反向传播算法（BackPropagation）是一种有监督神经网络学习算法，但原始算法收敛速率慢，训练过程易陷入局部极小值，精度不高等问题．文中提出了一种加权和引入参数改进的神经网络BP算法，某种程度上克服了以上缺点．对文中的改进算法用VC平台编程，并利用真实数据，对大学生就业能力进行了预测．实验表明，改进算法有效，也为高校解决大学生就业能力提供了决策支持．

Improvement of BP Algorithm Based on Weighted Mode

Back propagation （BP） neural network is a supervised neural network learning algorithm, but the original algorithm has a slow convergence rate and low accuracy, the training process of which is easier to fall into local minimum value. In this paper, we improved the BP neural network algorithm, which partily overcomes the above drawbacks. An improved BP algorithm is completed with the C programming language. We examined the improved BP algorithm with real data about the employment of students. As is shown that the modified algorithm is effective. And it is also a decision support algorithm for the ability of college students employment.