Locally linear embedding and neighborhood rough set-based gene selection for gene expression data classification
rough set (RS) are summarized. The advantages and disadvantages of these algorithms were analyzed and an optimized model for tumor gene selection was developed based on LLE and neighborhood

