Bayesian Unmasking Method for Positioning Multiple Blunder

Combining prior information with observing information,Bayesian methods for blunder detection are imposed.Especially a lot of effective measures are used to overcome the masking and swamping.When multiple blunder influence each other,the Bayesian method for blunder positioning based on the posterior probabilities of classification variables sometimes gives birth to masking and swamping which leads to the failure of positioning blunder.Hence,on the basis of seeking the reason of masking and swamping,and analyzing the eigenstructure of sampling correlation matrix of classification variables,the Bayesian unmasking method for positioning multiple blunder is introduced.The corresponding algorithm-adaptable MCMC sampling algorithm is implemented.