基于遗传算法和神经网络的矮小儿童智能诊断研究

针对儿童矮小问题,利用计算机辅助诊断方式建立基于神经网络的智能诊断系统。构造了矮小儿童评测的指标体系,利用遗传算法优化反向传播网络的结构参数,建立了一种基于反向传播网络的智能诊断模型。最后进行实例计算,结果表明该方法有效克服了纯反向传播网络算法局部收敛、泛化能力弱等问题,具有收敛速度快、精度高的优点,能有效适用于儿童矮小问题的智能诊断。

Research of intelligent diagnosis for short children based on genetic algorithms and neural networks

For the problem of short children, an intelligent diagnosis expert system is established based on neural networks. The index system of evaluation for short children is proposed. The back propagation network structure parameters are optimized by using genetic algorithms and a new intelligent diagnosis model is established by using back propagation neural network. The experiment results show that the new algorithm can overcome the local convergence and feebleness of generalization ability of back propagation algorithms. This method is superior to the other algorithms in the aspects of accuracy and accelerating convergence and is suitable for the computer-aided diagnosis of short children.