1 article
New methods reduce computational cost and handle constraints in optimization problems from PDEs to multi-agent reinforcement learning.