Taihei Oki

We present an efficient $O(n \ln T)$-regret method for online inverse linear optimization, extend it to suboptimal feedback, and provide an $\Omega(n)$-regret lower bound.