Feature selection algorithm based on P systems

Abstract Since the number of features dataset is much higher than patterns, dimension data, greater impact on learning algorithm. Dimension disaster has become an important problem. Feature selection can effectively reduce and improve performance Thus, in this paper, A feature algorithm based P systems (P-FS) proposed to exploit parallel ability cell-like advantage evolutionary algorithms search space select remove redundant information data. The P-FS tested five UCI datasets edible oil from practical applications. At same time, genetic (GAFS) are compared six datasets. experimental results show that good classification accuracy, stability, convergence. feasible selection.