Simulation And Computation Of An Exact Confidence Set For A Maximum Point Of A Univariate Polynomial Function In A Given Interval

In the article,Sanyu Zhou,Fang Wan,Wei Liu,Frank Bretz^[2]and Yang Han^[3]respectively,to study the change of market efficiency and the existence of a distorted trading model,and this paper is inspired by this article,discussed two methods of com-putation of an exact confidence set for a maximum point of a univariate polynomial function in a given interval.The conclusion of this paper is that the accuracy is based on the confidence bandwidth measured by the critical constant under the premise of ensuring the same accuracy requirements;The p-value method is better than the WL-HB method in terms of calculation speed,because the calculation of the critical value is the biggest difference between the two methods.The degree and speed of the prob-ability of finding the confidence band covering the maximum value point in the study is close to the given confidence level.Conclusion The p-value method is faster than the probability of the confidence band covering the maximum value point obtained by the WLHB method with the increase of the number of sample data sets given the con-fidence level and the number of simulations.The effect of the increased number of simulations on the WLHB method is more significant than that of the p-value method.When the given confidence level is 95%and 99%,that is,when the given confidence level is closer to 1,the effect of the number of simulations on the probability of the confidence band covering the maximum value point is gradually decreasing.The sim-ulation results show that the WLHB method is better when the number of sample data sets is less than the given confidence level and the number of simulation times.When the confidence level given by the combination is 90%,when the number of simulations is certain,the p-value method is better under this combination.