Reflection methods for user-friendly submodular optimization
arXiv:1311.4296
Abstract
Recently, it has become evident that submodularity naturally captures widely occurring concepts in machine learning, signal processing and computer vision. Consequently, there is need for efficient optimization procedures for submodular functions, especially for minimization problems. While general submodular minimization is challenging, we propose a new method that exploits existing decomposability of submodular functions. In contrast to previous approaches, our method is neither approximate, nor impractical, nor does it need any cumbersome parameter tuning. Moreover, it is easy to implement and parallelize. A key component of our method is a formulation of the discrete submodular minimization problem as a continuous best approximation problem that is solved through a sequence of reflections, and its solution can be easily thresholded to obtain an optimal discrete solution. This method solves both the continuous and discrete formulations of the problem, and therefore has applications in learning, inference, and reconstruction. In our experiments, we illustrate the benefits of our method on two image segmentation tasks.
Neural Information Processing Systems (NIPS), États-Unis (2013)
Cited by in corpus (10)
- Modular proximal optimization for multidimensional total-variation regularization
- Submodular Hypergraphs: p-Laplacians, Cheeger Inequalities and Spectral Clustering
- Parallel and Distributed Block-Coordinate Frank-Wolfe Algorithms
- Submodular Functions: from Discrete to Continous Domains
- Randomized Block Frank-Wolfe for Convergent Large-Scale Learning
- Revisiting Decomposable Submodular Function Minimization with Incidence Relations
- Active-set Methods for Submodular Minimization Problems
- Quadratic Decomposable Submodular Function Minimization: Theory and Practice (Computation and Analysis of PageRank over Hypergraphs)
- Graph Cuts with Interacting Edge Costs - Examples, Approximations, and Algorithms
- FPT Approximation Schemes for Maximizing Submodular Functions