Kriging methods for modeling spatial basis risk in weather index insurances: a technical note (Q6587515)

The authors fit a data set of weather phenomena on a cluster analysis model. The mathematical interest of this cluster analysis model is that it relies on the norms of the vectors for the specific data set. Weather insurance predictions are always significant to the insurance companies, because they may predict the probability of a non-desirable fact, which implies more payments for claims. The model of the authors may lead to a clearly non-parametric form of clustering, which is upon further research.