Indicator mining model for spatial multi-scale degraded land evaluation
Dianfeng Liu, Yaolin Liu, Xia Yin, Xiaofeng Hong, Zhongjun Zhao · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
At present, no feasible and effective methods meet the requirements of constructing a comprehensive and representative indicator system for degraded land evaluation, which orients spatial multi-scale and diversity of evaluation objects as well as integrates experts' judgments and objective information. This paper, tying to solve the problem, firstly proposes three propositions on evaluation indicator knowledge base (EIKB), universe evaluation indicator set (UEIS), evaluation indicator subset (EIS) and Mapping Rule of EIS (MREIS), and then constructs an heuristic indicator mining model (HIMM) based on above theories, variable precision rough set and information entropy. Finally, we applied HIMM to practical degraded land evaluation and examined the effectiveness of HIMM. The result shows that HIMM is applicable, especially in the aspect of solving the comprehensive and representative problem in the process of indicator system construction.