Accelerating Non-Negative and Bounded-Variable Linear Regression Algorithms with Safe Screening
This work addresses efficiency improvements for regression problems in machine learning and signal processing, but it is incremental as it adapts existing safe screening techniques to a new context.
The paper tackles the problem of accelerating solvers for non-negative and bounded-variable linear regression by identifying saturated coordinates during iterations, with experimental results showing compelling accelerations on synthetic and real data.
Non-negative and bounded-variable linear regression problems arise in a variety of applications in machine learning and signal processing. In this paper, we propose a technique to accelerate existing solvers for these problems by identifying saturated coordinates in the course of iterations. This is akin to safe screening techniques previously proposed for sparsity-regularized regression problems. The proposed strategy is provably safe as it provides theoretical guarantees that the identified coordinates are indeed saturated in the optimal solution. Experimental results on synthetic and real data show compelling accelerations for both non-negative and bounded-variable problems.