Abstract:We establish strong duality relations for functional two-step compositional risk-constrained learning problems with multiple nonconvex loss functions and/or learning constraints, regardless of nonconvexity and under a minimal set of technical assumptions. Our results in particular imply zero duality gaps within the class of problems under study, both extending and improving on the state of the art in (risk-neutral) constrained learning. More specifically, we consider risk objectives/constraints which involve real-valued convex and positively homogeneous risk measures admitting dual representations with bounded risk envelopes, generalizing expectations and including popular examples, such as the conditional value-at-risk (CVaR), the mean-absolute deviation (MAD), and more generally all real-valued coherent risk measures on integrable losses as special cases. Our results are based on recent advances in risk-constrained nonconvex programming in infinite dimensions, which rely on a remarkable new application of J. J. Uhl's convexity theorem, which is an extension of A. A. Lyapunov's convexity theorem for general, infinite dimensional Banach spaces. By specializing to the risk-neutral setting, we demonstrate, for the first time, that constrained classification and regression can be treated under a unifying lens, while dispensing certain restrictive assumptions enforced in the current literature, yielding a new state-of-the-art strong duality framework for nonconvex constrained learning.

CONJUGATE DUALITY FOR SET-VALUED VECTOR OPTIMIZATION IN FINITE DIMENSIONAL SPACES

An approach to duality via image space analysis and applications to linear vector optimization problems

Stable and Total Fenchel Duality for Convex Optimization Problems in Locally Convex Spaces

Duality and Optimization for Generalized Multi-Hop MIMO Amplify-and-Forward Relay Networks with Linear Constraints

STABLE LAGRANGE DUALITIES FOR ROBUST CONICAL PROGRAMMING

Dual Solutions in Convex Stochastic Optimization

Stable and Total Fenchel Duality for DC Optimization Problems in Locally Convex Spaces.

Variational Inequalities for Set-Valued Vector Fields on Riemannian Manifolds: Convexity of the Solution Set and the Proximal Point Algorithm.

Resolvents of Set-Valued Monotone Vector Fields in Hadamard Manifolds

Vector optimization problems with weakened convex and weakened affine constraints in linear topological spaces

Duality for a Class of Multiobjective Programing under Generalized V-Type I

Vector Variational-Like Inequalities with Constraints: Separation and Alternative

Optimality and duality analysis for multiobjective interval-valued semi-infinite optimization problem having vanishing constraints

Superefficiency in Vector Optimization with Nearly Subconvexlike Set-Valued Maps

Generalized Polyhedral Convex Optimization Problems

Lagrange Multipliers, Duality, and Sensitivity in Set-Valued Convex Programming via Pointed Closed Convex Processes

Dualities in Convex Algebraic Geometry

Strong Duality Relations in Nonconvex Risk-Constrained Learning

Some optimality conditions of set-valued optimization problems in locally convex topological vector spaces

Various duality models associated with some constrained robust nonlinear optimal control problems

Optimality conditions and duality for mathematical programming with equilibrium constraints including multiple interval-valued objective functions on Hadamard manifolds