In this paper we propose a
changepoint detection procedure based on a skew normal distribution from the
view of point of model selection. The detection procedure is constructed based
on Schwarz information criterion (SIC) combined with the binary segmentation
method for multiple changepoints detection purpose. Simulations are conducted
to illustrate the performance of the proposed test. We apply the method to detect
change points in the array Comparative Genomic Hy- bridization (aCGH) data set.