Functional Constrained Optimization for Risk Aversion and Sparsity Control
This addresses optimization challenges in applications like portfolio selection and healthcare planning, offering incremental improvements in projection-free methods for specific constraints.
The paper tackles functional constrained optimization problems with risk aversion and sparsity control, proposing projection-free methods like LCG, IPP-LCG, and DNCG that achieve iteration complexities such as O(1/ε^2 log(1/ε)) for convex cases and O(1/ε^3 log(1/ε)) or O(1/ε^4) for nonconvex cases.
Risk and sparsity requirements often need to be enforced simultaneously in many applications, e.g., in portfolio optimization, assortment planning, and treatment planning. Properly balancing these potentially conflicting requirements entails the formulation of functional constrained optimization with either convex or nonconvex objectives. In this paper, we focus on projection-free methods that can generate a sparse trajectory for solving these challenging functional constrained optimization problems. Specifically, for the convex setting, we propose a Level Conditional Gradient (LCG) method, which leverages a level-set framework to update the approximation of the optimal value and an inner conditional gradient oracle (CGO) for solving mini-max subproblems. We show that the method achieves $\mathcal{O}\big(\frac{1}{ε^2}\log\frac{1}ε\big)$ iteration complexity for solving both smooth and nonsmooth cases without dependency on a possibly large size of optimal dual Lagrange multiplier. For the nonconvex setting, we introduce the Level Inexact Proximal Point (IPP-LCG) method and the Direct Nonconvex Conditional Gradient (DNCG) method. The first approach taps into the advantage of LCG by transforming the problem into a series of convex subproblems and exhibits an $\mathcal{O}\big(\frac{1}{ε^3}\log\frac{1}ε\big)$ iteration complexity for finding an ($ε,ε$)-KKT point. The DNCG is the first single-loop projection-free method, with iteration complexity bounded by $\mathcal{O}\big(1/ε^4\big)$ for computing a so-called $ε$-Wolfe point. We demonstrate the effectiveness of LCG, IPP-LCG and DNCG by devising formulations and conducting numerical experiments on two risk averse sparse optimization applications: a portfolio selection problem with and without cardinality requirement, and a radiation therapy planning problem in healthcare.