中文说明：本程序通过Kmeans聚类算法对事件进行聚类，并在此基础上，通过神经网络来预测新输入案例可能发生的结果进行预测。由于隐私要求，程序的训练数据没有包含在内，但是作为大的框架来说，结构比较完整。
English Description：
This program uses kmeans clustering algorithm to cluster events, and on this basis, uses neural network to predict the possible results of new input cases. Due to privacy requirements, the training data of the program is not included, but as a large framework, the structure is relatively complete.