The nearest neighbor method is a simple and flexiblescheme for the classification of data points in a vector space. It predictsa class label of an unseen data point using a majority rule for the labels ofknown data points inside a neighborhood of the unseen data point. Becauseit sometimes achieves good performance even for complicated problems,several derivatives of it have been studied. Among them, the discriminantadaptive nearest neighbor method is particularly worth revisiting to demon-strate its application. The main idea of this method is to adjust the neigh-bor metric of an unseen data point to the set of known data points beforelabel prediction. It often improves the prediction, provided the neighbormetric is adjusted well. For statistical shape analysis, shape classificationattracts attention because it is a vital topic in shape analysis. However, be-cause a shape is generally expressed as a matrix, it is non-trivial to applythe discriminant adaptive nearest neighbor method to shape classification.Thus, in this study, we develop the discriminant adaptive nearest neighbormethod to make it slightly more useful in shape classification. To achievethis development, a mixture model and optimization algorithm for shapeclustering are incorporated into the method. Furthermore, we describe sev-eral helpful techniques for the initial guess of the model parameters in theoptimization algorithm. Using several shape datasets, we demonstrated thatour method is successful for shape classification.