Fast computing for dynamic screening systems when analyzing correlated data

ABSTRACT

In practice, we often need to identify individuals whose longitudinal behaviour is different from the behaviour of those well-functioning individuals, so that some unpleasant consequences (e.g. stroke) can be avoided or early detected. To handle such applications, a new statistical method, called dynamic screening system, has been developed in the literature. A recent version of this method can analyze correlated data. However, the computation involved is intensive. In this paper, we suggest a fast computing algorithm for the dynamic screening system. The algorithm can improve the effectiveness of the conventional dynamic screening system in certain cases. Numerical results show that the new algorithm works well in different cases.

KEYWORDS:

• Computing time
• correlation
• data de-correlation
• sequential monitoring
• sprint length
• statistical process control

Acknowledgments

The authors thank a referee for a number of helpful comments and suggestions. This research is supported in part by an NSF grant.

Disclosure statement

No potential conflict of interest was reported by the authors.