Optimal Mistake Bounds for Transductive Online Learning | Read Paper on Bytez