A Changepoint Detection Method for Profile Variance

A wavelet-based changepoint method is proposed that determines when the variability of the noise in a sequence of functional profiles goes out-of-control from a known, fixed value. The functional portion of the profiles are allowed to come from a large class of functions and may vary from profile to profile. The proposed method makes use of the orthogonal properties of wavelet projections to accurately and efficiently monitor the level of noise from one profile to the next. Several alternative implementations of the estimator are compared on a variety of conditions, including allowing the wavelet noise subspace to be substantially contaminated by the profile’s functional structure. The proposed method is shown to be very efficient at detecting when the variability has changed through an extensive simulation study.