Effective Distance based Low-Rank Preserving Projection

Preserving global and local structures during projection learning is very important for feature extraction, they commonly introduce an extra graph regularization term, If this initial graph construction is of low quality then the resulting projection may also be of low quality. Moreover, conventional projection methods simply use Euclidean distance to measure the similarity of samples, ignoring the dynamic structure information of data. Meanwhile, recent studies have shown that a probabilistically motivated distance measurement (called effective distance) can model the dynamic structure information of data. In this paper, we propose an effective distance based Low-Rank Preserving Projection method (EDLRPP), EDLRPP imposes the effective distance based graph constraint on the reconstruction error of data instead of introducing the extra regularization term to capture the global and local structure of data, which can greatly reduce the complexity of the model. We evaluate our method on datasets in the task of image classification, with results demonstrating the effectiveness of the method.