TY - RPRT TI - An Inexact Trust-Region Method for Structured Nonsmooth Optimization with Application to Risk-Averse Stochastic Programming AU - Drew P. Kouri PY - 2026 UR - https://arxiv.org/abs/2604.07216 ID - 2604.07216 ER -