提出了计算几何应用到时间序列挖掘的方法，实现了时间序列全序列匹配查询、模式查询、反向查询和异常检测，查询效率和准确性都有了比较大的提高。
By making use of the proximity query method in computational geometry, the whole matching query, pattern query, inverse query and outlier detection in time series are studied.
选取正交设计的试验点作为反向传播人工神经网络的训练集，实现对全实验域试验点的预测，并与实测的试验数据比较。
The trials of orthogonal design were chosen as the training set of the BP-ANN, and forecasted the trails of the entire field.
应用推荐