A Unified Algorithmic Framework for Block-Structured Optimization Involving Big Data: With applications in machine learning and signal processing
A Unified Algorithmic Framework for Block-Structured Optimization Involving Big Data: With applications in machine learning and signal processing
复制标题
DOI：
10.1109/msp.2015.2481563
复制
发表时间：
2015-11
影响因子：
14.9
通讯作者：
Mingyi Hong;Meisam Razaviyayn;Z. Luo;J. Pang
中科院分区：
文献类型：
--
作者：
Mingyi Hong;Meisam Razaviyayn;Z. Luo;J. Pang
本文提出了一个强大的大数据优化算法框架，称为块连续上限最小化（BSUM）。BSUM包括许多用于分析海量数据集的著名方法作为特例，例如块坐标下降（BCD）方法、凹凸过程（CCCP）方法、块坐标邻近梯度（BCPG）方法、非负矩阵分解（NMF）方法、期望最大化（EM）方法等。从设计灵活性、计算效率、并行/分布式实现以及所需的通信开销的角度讨论了BSUM的各种特征和属性。从网络，信号处理和机器学习的说明性的例子来展示BSUM框架的实际性能。
This article presents a powerful algorithmic framework for big data optimization, called the block successive upper-bound minimization (BSUM). The BSUM includes as special cases many well-known methods for analyzing massive data sets, such as the block coordinate descent (BCD) method, the convex-concave procedure (CCCP) method, the block coordinate proximal gradient (BCPG) method, the nonnegative matrix factorization (NMF) method, the expectation maximization (EM) method, etc. In this article, various features and properties of the BSUM are discussed from the viewpoint of design flexibility, computational efficiency, parallel/distributed implementation, and the required communication overhead. Illustrative examples from networking, signal processing, and machine learning are presented to demonstrate the practical performance of the BSUM framework.