Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore the latest advancements in continuous nonlinear nonconvex optimization solvers within the JuliaSmoothOptimizers (JSO) organization in this 17-minute presentation. Gain insights into why these solvers are suitable for large-scale optimization problems and learn about new packages like AdaptiveRegularization.jl. Discover how research-level solvers leverage high-level linear algebra packages, utilize flexible modeling strategies with Automatic Differentiation (AD) or tools like JuMP, and employ focused implementations. Understand the benefits of in-place solvers, multiprecision solves, and GPU computations. Examine the application of factorization-free solvers in addressing large-scale optimization problems, such as those arising from discretized PDE-constrained optimization. Learn about the JSO ecosystem's organization across various packages and how JSOSuite.jl serves as a comprehensive entry point for first-time users, simplifying benchmarking and introducing automatic algorithm selection. Explore the longevity and widespread adoption of JuliaSmoothOptimizers since 2015, with over 50 registered packages contributing to linear/nonlinear optimization and linear algebra domains across diverse ecosystems.
