Contingency-Aware Influence Maximization: A Reinforcement Learning Approach