Finding the Stochastic Shortest Path with Low Regret:
            The Adversarial Cost and Unknown Transition Case

                         Liyu Chen 1 Haipeng Luo 1

             Abstract                 end within a fixed number of steps is extensively studied
    We make significant progress toward the stochas-      in recent years (often known as episodic finite-horizon re-
    tic shortest path problem with adversarial costs      inforcement learning or loop-free SSP). The general (and
    ...