Predicting ad click-through rates is the core problem in display advertising, which has received much attention from the machine learning community in recent years. In this paper, we present an online learning algorithm for click-though rate prediction, namely Follow-The-Regularized-Factorized-Leader (FTRFL), which incorporates the Follow-The-Regularized-Leader (FTRL-Proximal) algorithm with per-coordinate learning rates into Factorization machines. Experiments on a real-world advertising dataset show that the FTRFL method outperforms the baseline with stochastic gradient descent, and has a faster rate of convergence.