An efficient algorithm for a class of fused lasso problems
Authors:
Open Access
- 25 July 2010
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 323-332
Abstract
The fused Lasso penalty enforces sparsity in both the coefficients and their successive differences, which is desirable for applications with features ordered in some meaningful way. The resulting problem is, however, challenging to solve, as the fused Lasso penalty is both non-smooth and non-separable. Existing algorithms have high computational complexity and do not scale to large-size problems. In this paper, we propose an Efficient Fused Lasso Algorithm (EFLA) for optimizing this class of problems. One key building block in the proposed EFLA is the Fused Lasso Signal Approximator (FLSA). To efficiently solve FLSA, we propose to reformulate it as the problem of finding an "appropriate" subgradient of the fused penalty at the minimizer, and develop a Subgradient Finding Algorithm (SFA). We further design a restart technique to accelerate the convergence of SFA, by exploiting the special "structures" of both the original and the reformulated FLSA problems. Our empirical evaluations show that, both SFA and EFLA significantly outperform existing solvers. We also demonstrate several applications of the fused Lasso.Keywords
This publication has 23 references indexed in Scilit:
- Properties and refinements of the fused lassoThe Annals of Statistics, 2009
- Recovering time-varying networks of dependencies in social and biological studiesProceedings of the National Academy of Sciences, 2009
- Classification of arrayCGH data using fused SVMBioinformatics, 2008
- A coordinate gradient descent method for nonsmooth separable minimizationMathematical Programming, 2007
- Primal-dual subgradient methods for convex problemsMathematical Programming, 2007
- Regional copy number–independent deregulation of transcription in cancerNature Genetics, 2006
- Minimizing Quadratic Functions Subject to Bound Constraints with the Rate of Convergence and Finite TerminationComputational Optimization and Applications, 2005
- Sparsity and Smoothness Via the Fused LassoJournal of the Royal Statistical Society Series B: Statistical Methodology, 2004
- An algorithm for the solution of certain tridiagonal systems of linear equationsThe Computer Journal, 1972
- An algorithm for solving a special class of tridiagonal systems of linear equationsCommunications of the ACM, 1969