Maximizing Correlation in the Presence of Missing Data
Xinfang Wang · Applied mathematical sciences · 2008
In this paper we address the problem of maximizing the correlation between two vectors of time series data, when one of the vectors has missing data and the timing of the missing data is unknown. The motivation for this work comes from environmental monitoring where because of monitoring malfunction, some data are lost. We study the use of integer programming and a genetic algorithm (GA) for this problem.