A genetic algorithm for Hierarchical Multi-Label Classification

摘要: In Hierarchical Multi-Label Classification (HMC) problems, each example can be classified into two or more classes simultaneously, differently from standard classification. Moreover, the are structured in a hierarchy, form of either tree directed acyclic graph. Therefore, an assigned to paths hierarchical structure, resulting complex classification problem with possibly hundreds thousands classes. Several methods have been proposed deal such some them employing single classifier all simultaneously (global methods), and others many classifiers decompose original set subproblems (local methods). this work, we propose novel global method called HMC-GA, which employs genetic algorithm for solving HMC problem. our approach, evolves antecedents rules, order optimize level coverage antecedent. Then, optimized is selected build corresponding consequent rules (set predicted). Our compared state-of-the-art algorithms, protein function prediction datasets. The experimental results show that approach presents competitive predictive accuracy, suggesting algorithms constitute promising alternative multi-label biological data.