An Alternating Direction Method of Multipliers Algorithm for the Weighted Fused LASSO Signal Approximator
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913445821022208 |
|---|---|
| author | Dijkstra, Louis Hanke, Moritz Koenen, Niklas Foraita, Ronja |
| author_facet | Dijkstra, Louis Hanke, Moritz Koenen, Niklas Foraita, Ronja |
| contents | We present an Alternating Direction Method of Multipliers (ADMM) algorithm designed to solve the Weighted Generalized Fused LASSO Signal Approximator (wFLSA). First, we show that wFLSAs can always be reformulated as a Generalized LASSO problem. With the availability of algorithms tailored to the Generalized LASSO, the issue appears to be, in principle, resolved. However, the computational complexity of these algorithms is high, with a time complexity of $O(p^4)$ for a single iteration, where $p$ represents the number of coefficients. To overcome this limitation, we propose an ADMM algorithm specifically tailored for wFLSA-equivalent problems, significantly reducing the complexity to $O(p^2)$. Our algorithm is publicly accessible through the R package wflsa. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_18077 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | An Alternating Direction Method of Multipliers Algorithm for the Weighted Fused LASSO Signal Approximator Dijkstra, Louis Hanke, Moritz Koenen, Niklas Foraita, Ronja Methodology We present an Alternating Direction Method of Multipliers (ADMM) algorithm designed to solve the Weighted Generalized Fused LASSO Signal Approximator (wFLSA). First, we show that wFLSAs can always be reformulated as a Generalized LASSO problem. With the availability of algorithms tailored to the Generalized LASSO, the issue appears to be, in principle, resolved. However, the computational complexity of these algorithms is high, with a time complexity of $O(p^4)$ for a single iteration, where $p$ represents the number of coefficients. To overcome this limitation, we propose an ADMM algorithm specifically tailored for wFLSA-equivalent problems, significantly reducing the complexity to $O(p^2)$. Our algorithm is publicly accessible through the R package wflsa. |
| title | An Alternating Direction Method of Multipliers Algorithm for the Weighted Fused LASSO Signal Approximator |
| topic | Methodology |
| url | https://arxiv.org/abs/2407.18077 |