同一个记录的参数相互不匹配，可用神经网络解决时差问题；两类样本数目悬殊

The parameters of the same record do not match each other, and the neural network can be used to solve the problem of jet lag; the number of samples of the two types is very different. For example, the number of accidents is very small, and it is difficult to achieve the purpose of training. You can find another one that is related to the occurrence of accidents and appears more frequently. The process parameters are taken as the specific target; the optimization result is of no value, which may be caused by the unlimited optimization conditions; the optimization target is hidden in the variable, and the variable should be used as the optimization target at this time.

If the parameters of the same record do not match each other, the time difference problem can be solved by neural network; There is a great difference in the number of samples between the two types. For example, there are few samples of accidents, so it is difficult to achieve the purpose of training. Another process parameter related to accidents and occurring frequently can be found as the specific target; The optimization result has no value, which may be caused by the unlimited optimization conditions; The optimization objective is hidden in the variable, which should be taken as the optimization objective.

The parameters of the same record do not match each other, so the time difference problem can be solved by neural network. The number of the two types of samples is very different. For example, there are few samples of accidents, so it is difficult to achieve the purpose of training. We can find another process parameter that is related to accidents and appears frequently as a specific goal. The result of optimization has no value, which may be caused by unlimited optimization conditions; The optimization target is hidden in the variable, so this variable should be taken as the optimization target.