Study And Comparison On The Index Weighting Approach In Land Resource Evaluation

This paper take the rural areas land consolidation evaluation of Pingliang Provincezhuanglang County for example, described a variety of empowering methods in detail and dosome case studies on the empowering methods. First of all, introduce the land resourceevaluation’s concept and meaning, described the status and development of land resourceassessment from production potential, suitability, economic, sustainable use, ecological securityand intensive use six areas; besides to category description the application of empower methodsin the field of land resources evaluation.In the second, combined with Zhuanglang county ruralresidential consolidation potential evaluation, select the six objective weighting methods, such asentropy method, maximizing deviation method, variation coefficient method, mean squaredeviation method, factor analysis and gray relation method, from the two aspects of reduce theindex data and sample data, analyzed and compared the changes of the weighting results. Finally,concluding the paper and pointed out the deficiencies of this article and to work towards.Through case studies, the results showed that:1)Entropy method in the index weighting process whatever the index system changes or sampledata changes,are highlighted on the sensitivity of the index difference, can be clearly distinctionbetween indicators. According to significant sensitivity of the index change, determine theentropy method is the bests index weighting method in these methods.2)Maximizing deviation method, mean square deviation method, variation coefficient methodand entropy method is very similar essentially, the relative degree of dispersion of the evaluationare based on each index object attribute values to determine the weight of each index size. Whichmaximizing deviation method and mean square deviation method in the case of small sampledata indicators poor distinction,so that can not be clearly separated from each index. Thevariation coefficient method can be a clear distinction between the individual index, but thediscrimination is less than the entropy method.3)Factor analysis and gray relation method due to the defects of the method itself, the weaksensitivity of the index difference, not be suitable for the empowerment of land evaluation. Andeach weighting method for its scope of application, such as factor analysis has a high correlationbetween the indexes. Factor analysis requires evaluation in the case of have a prerequisite, not abuse.4) When a variety of methods of evaluation differences distinguish or less, which means that theresults of various methods of approximation, you can select any index weighting methodempowering; when the distinction of the various methods on the index difference time, theproposed entropy method.