FILTER-BASED CUCKOO OPTIMIZATION ALGORITHM FOR FEATURE SELECTION

Majority of the stored data contain irrelevant and redundant features that affect the classification performance. Therefore, feature selection (FS) is essential to remove both the redundant and ir- relevant features and consequently improved the classification performance. FS can be either fil- ter or wrapper. In wrapper FS, a classifier is used to measure the accuracy of the selected fea- tures but its computationally expensive and not favourable on large dimensional datasets.