Joint of locality- and globality-preserving projections

Dimensionality reduction is an important topic in machine learning community, which is widely used in the areas of face recognition, visual detection and tracking. Preserving local and global structures simultaneously is crucial for dimensionality reduction. In this paper, local and global approaches are generalized, respectively, and then a unified framework that joins the effective local and global terms is presented for unsupervised dimensionality reduction. Furthermore, to search for the optimal integration parameter, the proposed method uses two different search schemes named JLGP and IJLGP, respectively, where JLGP corresponds to the manual search scheme and IJLGP corresponds to the automatic search schemes. The promising experimental results on four benchmark datasets validate the effectiveness of the proposed method.