Name
Abstract | ||
|---|---|---|
In this paper we study the estimation of changing trends in time-series using l(1) trend filtering. This method generalizes 1D Total Variation (TV) denoising for detection of step changes in means to detecting changes in trends, and it relies on a convex optimization problem for which there are very efficient numerical algorithms. It is known that TV denoising suffers from the so-called stair-case effect, which leads to detecting false change points. The objective of this paper is to show that l(1) trend filtering also suffers from a certain staircase problem. The analysis is based on an interpretation of the dual variables of the optimization problem in the method as integrated random walk. We discuss consistency conditions for l(1) trend filtering, how to monitor their fulfillment, and how to modify the algorithm to avoid the stair-case false detection problem. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1109/ICASSP.2015.7178711 | 2015 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING (ICASSP) |
Keywords | Field | DocType |
l(1) trend filtering, generalized lasso, TV denoising, Fused Lasso, change point detection | Noise reduction,False detection,Mathematical optimization,Random walk,Computer science,Casing,Computational mathematics,Filter (signal processing),Algorithm,Convex optimization,Optimization problem | Conference |
Volume | ISSN | Citations |
abs/1412.0607 | 1520-6149 | 1 |
PageRank | References | Authors |
0.35 | 2 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Cristian R. Rojas | 1 | 252 | 43.97 |
| Bo Wahlberg | 2 | 210 | 40.68 |