ResHDrch of classification for defective components of automotive recall based on clustering algorithm

The classification of Recall defective components in China is unsuitable for further research and facial operation. Hierarchical clustering algorithm was present to cluster the classification of vehicle parts and components which will be used in vehicle defects recall. Based on QC/T 265-2004 named regulation of vehicle parts and components and American NHTSA recall data, 32 classes was clustered through digital processing of the attribute value of every dimension of data samples that can distinguish one component from the other. Obviously the clustered results are suitable for facial operation and reasonable, meanwhile they partly avoided from unreasonable classes.