As a kind of model-based optimization framework, the sampling-and-classification (SAC) algorithms, where the model is specified to be a classifier, has been recently studied in both theoretical foundation and algorithm implementation. However, the previous work only studied SAC algorithms in real do