A Cuckoo Search Optimization based Method for Feature Selection

Today, Feature Selection, as a technique to improve classification methods, has attracted the attention of many computer scientists. As the methods of extracting data is improving every day, we encounter matrices with huge dimension and it will affect the performance of processing. Hence, reducing the number of features by choosing the best subset of all features, will affect the algorithms performance. Finding the best subset by comparing all possible subsets, is an intractable process, as a solution, heuristic methods can help to find the near-optimal solutions with an acceptable complexity. In this paper, we introduce a novel Feature Selection technique which selects the most informative features and omits the redundant or irrelevant ones. Our method is embedded in CS (Cuckoo Search). The performance of our method is evaluated on two classification benchmarks: Vowel, and Wine. Comparing the results with four state-of-the-art methods, demonstrates its superiority over them.

Keywords:

Feature selection , Cuckoo Search , Classification , Machine Learning