Detecting and Disposing Abnormal Signal Outliers with Masking Effect by Using Data Accumulated Generating Operation

In this paper, we study a signal processing problem which concentrates on outlier detection and data mining in order to rediscover some useful information. Moreover, the great difficulty of the subject is caused by both the external environment and the internal mechanism. In the external, owing to limit of impersonal condition, such as, realistic disturbance from noise signal can not avoid. In the internal, the masking effect appears in outlying observation so as to produce some unreasonable results by data analysis. Therefore, based not only on the requirement of external quality assurance schemes, but also on internal quality control where screening for outliers should probably be part of the procedure for our main goal. As a consequence, we have applied data accumulated generating operation and sample median test for solving the conundrum.