deterministic case

Next, we extended the deterministic model to stochastic case. Again, an algorithm was brought up and some properties were studied.

进一步地，本文将前面的确定性模型扩展到随机的情况，并提出了算法和一些性质。

In this case, deterministic method can not be achieved, but the artificial neural networks are non-deterministic method can play a role.

这种情况下，确定性的分析方法很难奏效，而人工神经网络等非确定性方法却可以发挥作用。

Our results show that the error threshold should locate in a certain range for any limited sequence length instead of a single point, and the upper bound of the error threshold shifts to a larger value compared with the deterministic case.

结果表明对于有限长度的序列，误差阈应处于一定的范围之内，与确定情形相比其上限向较大值有一定平移。

In case of deterministic system, they are much closed and the present results are even better.

在确定系统, 其结果相当接近甚至更佳.

Using the ECMWF ensemble forecasting system as an example, a deterministic forecast case and a probability forecast case are discussed, respectively.

以欧洲中期天气预报中心的集合预报系统的控制预报和集合预报为例，对确定性预报和概率预报的情况分别进行了说明。