2021/06/16 by Qing Ye, Weijun Xie, Ye, Qing +1 · 1 citation
Decision Sciences · Engineering · Mathematics · #Advanced Optimization Algorithms Research #Algorithm #Applied mathematics #Computer science #Cone (formal languages) #Conic section #Exponential function #FOS: Mathematics #Geometry #Integer (computer science) #Logarithm #Mathematical analysis #Mathematics #Optimization and Control (math.OC) #Optimization and Mathematical Programming #Risk and Portfolio Optimization #Scaling #math.OC
paper · pdf · doi:10.48550/arxiv.2106.09123
37 pages, 9 figures
openalex publication_date 2021/06/16 · arxiv created 2022/03/19 · arxiv updated 2022/03/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Exponents and logarithms are fundamental components in many important applications such as logistic regression, maximum likelihood, relative entropy, and so on. Since the exponential cone can be viewed as the epigraph of perspective of the natural exponential function or the hypograph of perspective of the natural logarithm function, many mixed-integer convex programs involving exponential or logarithm functions can be recast as mixed-integer exponential conic programs (MIECPs). However, unlike mixed-integer linear programs (MILPs) and mixed-integer second-order conic programs (MISOCPs), MIECPs are still under development. To harvest the past efforts on MILPs and MISOCPs, this paper presents second-order conic (SOC) and polyhedral approximation schemes for the exponential cone with application to MIECPs. To do so, we first extend and generalize existing SOC approximation approaches in the extended space, propose new scaling and shifting methods, prove approximation accuracies, and derive lower bounds of approximations. We then study the polyhedral outer approximation of the exponential cones in the original space using gradient inequalities, show its approximation accuracy, and derive a lower bound of the approximation. When implementing SOC approximations, we suggest learning the approximation pattern by testing smaller cases and then applying it to the large-scale ones; and for the polyhedral approximation, we suggest using the branch and cut method for MIECPs. Our numerical study shows that the proposed methods show speed-ups over solver MOSEK for MIECPs, and the scaling, shifting, and polyhedral outer approximation methods work very well.