The Reselection of Adjacent Sets by Consistency-Based Feature Selection Algorithm
Sho Shimamura, Kouichi Hirata · 2019
In this paper, we introduce the reselection of adjacent sets by the consistency-based feature selection algorithm. Here, an adjacent set is the feature set obtained by applying the algorithm to the data set such that a single feature in the feature set selected by the algorithm is removed and we call such an application the reselection of adjacent sets. Then, we give a simple example that, for the algorithm CWC, the cardinality of an adjacent set is not always larger than that of the feature set selected by CWC. Next, we give experimental results to evaluate the results of the reselection of adjacent sets by CWC, by using real data and nucleotide sequences.