基于多维尺度变换和自适应增强回归进行插值 ———以三维地质属性体为例.

In order to explore the distribution characteristics of geological attributes, obtain the attribute values of unknown points in the 3D geological body and realize automatic interpolation, the spatial interpolation of the 3D geological attribute body was proposed by using multidimensional scaling (MDS) and adaptive boosting regression (ABR). In spatial interpolation, the attribute value of unknown points was closely related to the attribute value, distance, and anisotropy of its nearest sample points. MDS was used to reduce the dimension of these factors to obtain one-dimensional factors. Then the ABR interpolation model was established to obtain unknown points. finally, MF was used to reduce the discontinuity which was caused by ABR, and automatic interpolation was realized. By comparing the Ordinary Kriging method, the results show that this method simplified the cumbersome process of artificial semi-variation analysis in the Kriging method and realize automatic interpolation. at the same time, it avoid the limitation that the known sample points must comply with the specific random field when using Kriging interpolation in the geological zone of the non-specific random field. [ABSTRACT FROM AUTHOR]