Bootstrap法对时间序列问题预测区间的修正

Improved prediction for time series analysis problem using Bootstrap method

针对时间序列问题中线性预测的局限性,引入Bootstrap算法解决了MA模型和ARCH模型的预测问题.使得上述问题归结为一个条件期望值的求解,并对该预测问题进行了优化.

Due to the limitation of linear prediction,the Bootstrap algorithm was introduced to solve forecast question for MA model and ARCH model.And the above question can be seen as a condition expected value and was predicted.