Author(s)

Siyao Chu, Qunli Geng, Zhanbo Liang

Corresponding Author

Siyao Chu

Abstract

Spatial variable co-estimation is crucial in engineering and fields like geology/mining. Traditional methods like Kriging have limitations with complex correlations. This paper studies a method for spatial attribute data co-estimation. It preprocesses data, divides it into grids, and interpolates. By comparing BN-NN and KRF, it finds KRF is more accurate and robust, and it effectively estimates variables with insufficient sampling. This research provides a more reliable tool for related fields, enabling more accurate spatial data analysis and prediction, which is of great significance for optimizing resource exploration, environmental monitoring, and urban planning, and promoting the development of these fields.

Keywords

Spatial variable co-estimation; Kriging interpolation; Kriging random forest; Estimation error