The optimization method is one of the core techniques to solve practical problems arising from science and engineering. With the rapid development of data science, comprehensive integration of massive data and optimization methods are required to make engineering and business decisions smarter, more agile, and more efficient. The data-driven optimization method has become an important tool in many fields including machine learning, image science, signal processing, and so on. Optimization over data-driven systems is a challenging task, which has attracted significant attention in the recent literature. The main purpose of this special issue is to reflect the latest advances in theory and algorithms for data-driven optimization methods. We solicited 17 high-quality papers on theory and algorithms for data-driven optimization. Topics of the special issue cover the fields of nonconvex optimization theory and analysis, stochastic optimization and applications, matrix and tensor computational methods and theory, numerical optimization methods in machine learning, and optimization algorithms and theory in image science, signal processing.
展开▼
