{"id":"arxiv/2609.15679","url":"https://arxiv.org/abs/2609.15679","tags":["math.OC","cs.LG"],"score":0,"title":"Projection-Free Multi-level Algorithms for Stochastic Constrained Compositional Optimization","domain":"math.OC","source":"arxiv","authors":["Wei Jiang","Sifan Yang","Wenhao Yang","Yibo Wang","Yuanyu Wan","Zechao Li","Lijun Zhang"],"reviews":0,"subject":"math.OC","summary":"This paper studies projection-free algorithms for stochastic constrained multi-level compositional optimization. In this context, the objective function is a nested composition of several smooth functions, and the decision set is closed and convex. Since projection onto the constraint set can be computationally expensive, we develop projection-free methods that rely on linear minimization oracles. For non-convex objectives, we propose variance-reduced projection-free algorithms and establish complexity guarantees under both the Frank-Wolfe gap and the gradient mapping criteria. We also…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-14","reactions":{}}