Granular Space-Based Feature Selection and Its Applications

Feature selection is viewed as an important preprocessing step for pattern recognition, machine learning and data mining. Considering a consistency measure introduced in rough sets, the problem of feature selection aims to retain the discriminatory power of original features. Many heuristic feature selection algorithms have been proposed, however, these methods are computationally time-consuming. This paper introduces granular space, positive granular space and negative granular space based on granular computing in simplified decision systems, and then new feature significance measure is proposed. Meanwhile, their important propositions and properties are derived. Furthermore, by virtue of radix sorting and Hash techniques, the object granules as basic processing elements are employed to investigate feature selection, and then a heuristic algorithm with low computational complexity is explored. Numerical simulation experiments show that the proposed approach is indeed efficient, and therefore of practical value to many real-world problems.