Academic

I will discuss detection and estimation of local signals, represented by changes in a sequence of observations. Changes can occur continuously or discontinuously. Two conceptually similar, but technically different problems are (i) segmentation of independent normal observations according to abrupt changes in the mean and detection of changes in the slope of a regression line. Theoretical results will be illustrated by applications to copy number variation in the BT474 cell line and to historical weather/rainfall patterns. Confidence regions for the change-points will be discussed.