A Computational Model for Feature Selection Techniques /
نموذج محسب لتقنيات اختيار الدلالات
Sanaz Abdelfattah Sayed Gomaa ; Supervised Amr Badr, Emad Nabil
- Cairo : Sanaz Abdelfattah Sayed Gomaa , 2016
- 84 Leaves : charts ; 30cm
Thesis (M.Sc.) - Cairo University -Faculty of Computers and Information - Department of Computer Science
Many applications depend on large datasets with a lot of features, some of these features may be considered irrelevant, high dimensional or noisy that will degrade the perfor- mance of the machine learning tasks so, these applications use feature selection task as an important step in their implementation such as, data mining, classication, pattern recognition, and optimization. This task can be extremely useful in reducing the dimen- sional data to be processed by the classier, reducing the execution time and enhancing the recognition rate of the classier. Until now, nding the most informative data among the large data still an open prob- lem. For the feature selection problem, the goal is to search about the most informative subset of features that represent the original features in a specic domain. The selected features are used in optimization of a certain tness function, so the feature selection problem can be seen as an optimization problem. In the last years, the clonal selection was used to solve many problems of different applications where, it has an important role in the Articial Immune System (AIS) that describes an adaptive immune response to the stimulation of non-self-cells (antigens). This thesis presents two techniques to solve the feature selection issue, the rst one is an improved binary clonal selection algorithm (BCSA). While, the second is a new hybrid algorithm that combines Clonal Selection Algorithm (CSA) with Flower Pollination Algorithm (FPA) to compose Bi- nary Clonal Flower Pollination Algorithm (BCFA)