采用主成分分析的面状居民地匹配指标简化方法

现有多源居民地匹配中存在众多的面要素度量指标,若全部进行考虑,则增加了匹配的复杂性;若只考虑部分指标,则可能造成匹配信息的缺失,影响匹配结果。针对这一问题,本文提出一种采用主成分分析方法的面状居民地匹配方法。借鉴主成分分析法中降维的思想,对居民地各项度量指标进行定性定量分析,通过科学计算确定面要素匹配综合指标,用较少的新指标代替原来较多的相似性指标,进而根据获得的整体相似性评价指标进行居民地匹配。实验分析表明,本文方法简化了匹配过程中众多的相似性指标,降低了匹配复杂性和不确定性,避免了各相似权值确定较为随意的问题,有效提高了匹配效率和正确率。

A Habitation Matching Indicators Simplify Method by Using Principal Component Analysis

Existing habitation matching methods mostly use lots of matching indicators,it will increase the complexity of matching if consider all,but if consider part of the indicator,the matching information will miss and impact matching results.In response to these problems,A habitation matching indicators simplify method by using principal component analysis is proposed.Based on the principal component analysis idea of principal component analysis,analysis the habitation indicators metrics for qualitative and quantitative,through scientific calculation to determine comprehensive indicators of surface matching elements,with fewer new indicators instead of more original similar indicators,then according to obtain the overall similarity evaluation indicators of habitation to match.Experiments show that method simplifies the matching process in many similarity metrics,can reduce the matching complexity and uncertainty,to avoid the similar weights are free of problems,effectively improve the matching efficiency and accuracy.