Intelligent Optimization on Power Values for Inverse Distance Weighting

As a typical spatial interpolation method in geoscience and geographic data processing, inverse distance weighting (IDW) method has a long-time standing problem which is how to choose the best power value while calculating. For the issue, this text proposes an alternative way based on enumeration and cross validation. By orderly analyzing and estimating the effectiveness of every possible power value through cross-validation, this method can intelligently find the proper power value. Additionally, a practical application procedure with a real dataset by the proposed method is carried out as a typical case study, from which the calculation result is obviously superior to conventional method. Thus, it is not difficult to conclude that both theoretical analysis and practical application can show the validity and practicability of the proposed method.