When the deformation of the dam structure reaches a critical value, a possible dam failure event may occur. Therefore, dam deformation monitoring and accurate prediction are essential. Considering that previous dam deformation prediction models have problems such as low prediction accuracy and inability to meet actual engineering needs, this paper proposes a novel method based on support vector machines to solve this problem. Specifically, the Chameleon Optimization (CO) algorithm is proposed. The CO algorithm is used to optimize the penalty factor and kernel function parameters of the Support Vector Machine (SVM), and a dam deformation prediction model based on the CO-SVM algorithm is established. Particle Swarm Optimization (PSO) algorithm and Gray Wolf Optimization (GWO) algorithm were introduced to optimize SVM to construct the PSO-SVM model and GWO-SVM model respectively. Through the measured data of Shuibuya concrete-faced rockfill dam, the prediction results are compared with the prediction results of SVM, PSO-SVM and GWO-SVM models. The final experimental results verified the effectiveness and superiority of the method.