A Method for the Identification of Sample Classification Based on TCM-SVDD

The identification of the mechanical noise sources of a ship may be considered as a pattern recognition problem based on small samples,which can be effectively solved by applying the increment learning method. However,the new increment samples must be classified prior to the increment learning. Aiming at the problem,an innovative approach,named TCM-SVDD (Transductive Confidence Machine for Support Vector Data Description),is proposed in this paper for the classification of new samples. The strangeness of all samples are first calculated using the SVDD method. Then,the confidence degree of new samples are estimated and compared with the presettable confidence level. Finally,the identification of heterogeneous pattern samples is accomplished. The results show that the heterogeneous pattern samples can be rapidly identified with the proposed method,and it is applicable even when there are only few heterogeneous samples in the training sample set.