We discuss the methods and results of the RESSTE team in the competition on spatial statistics for large datasets. In the first sub-competition, we implemented block approaches both for the estimation of the covariance parameters and for prediction using ordinary kriging. In the second sub-competition, a two-stage procedure was adopted. In the first stage, the marginal distribution is estimated neglecting spatial dependence, either according to the flexible Tuckey g and h distribution or nonparametrically. In the second stage, estimation of the covariance parameters and prediction are performed using Kriging. Vecchias’s approximation implemented in the GpGp package proved to be very efficient. We then make some propositions for future competitions.

我们讨论了 RESSTE 团队在大型数据集空间统计竞赛中的方法和结果。在第一个子竞赛中，我们实现了块方法来估计协方差参数和使用普通克里金法进行预测。在第二次分赛中，采用了两阶段的程序。在第一阶段，根据灵活的 Tuckey g和h分布或非参数地估计边缘分布，忽略空间依赖性。在第二阶段，协方差参数的估计和预测是使用克里金法进行的。在GpGp包中实现的Vecchias近似被证明是非常有效的。然后我们为未来的比赛提出一些建议。