2019/12/19 by Ilnura Usmanova, Andreas Krause, Usmanova, Ilnura +3 · 1 citation
Computer Science · Decision Sciences · #Advanced Bandit Algorithms Research #FOS: Mathematics #Machine Learning and Algorithms #Numerical Analysis (math.NA) #Optimization and Control (math.OC) #Stochastic Gradient Optimization Techniques
paper · pdf · doi:10.48550/arxiv.1912.09466
openalex publication_date 2019/12/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
For safety-critical black-box optimization tasks, observations of the constraints and the objective are often noisy and available only for the feasible points. We propose an approach based on log barriers to find a local solution of a non-convex non-smooth black-box optimization problem min f0(x) subject to fi(x)≤ 0,~ i = 1,…, m, at the same time, guaranteeing constraint satisfaction while learning an optimal solution with high probability. Our proposed algorithm exploits noisy observations to iteratively improve on an initial safe point until convergence. We derive the convergence rate and prove safety of our algorithm. We demonstrate its performance in an application to an iterative control design problem.