Research on Classifier of MultiDimensional Data Based on Radar Chart Mapping and Fuzzy Rule Reasoning
Hui Meng, Wenxue Hong, Liqiang Wang, Jialin Song · 2006
A novel method for designing the classifier of multidimensional data was proposed, which used radar chart of multi-statistics to show multidimensional data and applied fuzzy rules to recognize the radar chart in order to implement automatic classification. Different multidimensional data formed different radar chart and distinguished different category. The treatment of radar chart gave prominence to tiny variable that conduced to the classification so that the information of all variables was fully used. Testing data of 132 compound oil samples suggests that this classifier may accurately distinguish all compound oils from each other except for several illogical samples and presents better result of category and precision