A New Supervised Discriminant Locality Preserving Projections Algorithm

When performing visualization and classification, people often confront the problem of dimensionality reduction. Recently, a new manifold learning algorithm named Locality Preserving Projections (LPP) that aims at finding an embedding that preserves local information has been proposed. But LPP is an unsupervised algorithm. Supervised locality preserving projection (SLPP) seeks to find the projection which efficiently preserves the local structure of data points embedded in high-dimensional data space. However, it has the over-learning problem and does not preserve the class and discriminative information of data which is also useful for data recognition. To solve the problem, a novel feature extraction method based oil discriminant information, namely supervised discriminative locality preserving projections (SDLPP), is presented in this paper. The SDLPP takes into account class and discriminative information of data points. Different from the most existing supervised locality preserving projection methods, the SDLPP not only preserves both the local structure and diversity information of data, but also avoids the data over-fitting problem. We compare the proposed SDLPP with LPP and SLPP methods on different data sets. Experimental results suggest the efficiency of the proposed method.