Cross-validation is a set of techniques used to evaluate how accurate a predictive model is in predicting a set of independent data.