Detecting high dimensional correlated data streams is becoming more and more popular for real alarming time out of control in many practical application. But in order to finding the stopping time as soon as possible a...Detecting high dimensional correlated data streams is becoming more and more popular for real alarming time out of control in many practical application. But in order to finding the stopping time as soon as possible after the drift occurred in the activity, we should choose appropriate control chart. This article compared the ARL＇s of the CUSUM and EWMA charts out of control based on same average run length in control Calculating the sum and max of CUSUM and EWMA statistics respectively, giving appropriate control limit for control charts, and comparing the balance between robustness and sensitivity.展开更多
文摘Detecting high dimensional correlated data streams is becoming more and more popular for real alarming time out of control in many practical application. But in order to finding the stopping time as soon as possible after the drift occurred in the activity, we should choose appropriate control chart. This article compared the ARL＇s of the CUSUM and EWMA charts out of control based on same average run length in control Calculating the sum and max of CUSUM and EWMA statistics respectively, giving appropriate control limit for control charts, and comparing the balance between robustness and sensitivity.
