分别采用BP网络和RBF网络进行数据预测
在600085这个数据表中，以XZSLX综合作为输入，以收盘价作为输出
以前595个数据作为训练样本，后100个数据作为实际输出。
通过训练数据建立模型，最终输出100个预测的股票收盘价。
将预测的收盘价和实际的收盘价进行对比并求取误差，从而判断所建立模型的准确性。
BP神经网络调用newff函数实现。(In the data table of 600085, XZSLX synthesis is used as input and closing price as output.

The former 595 data are used as training samples, and the latter 100 data are used as actual output.

Through training data to build a model, the final output of 100 predicted stock closing prices.

By comparing the predicted closing price with the actual closing price and calculating the error, the accuracy of the established model can be judged.

BP neural network is implemented by calling newff function.)