A grammatical evolution algorithm for generation of Hierarchical Multi-Label Classification rules

摘要: Hierarchical Multi-Label Classification (HMC) is a challenging task in data mining and machine learning. Each instance HMC can be classified into two or more classes simultaneously. These are structured hierarchy, the form of either tree directed acyclic graph. Therefore, an assigned to paths from hierarchical structure, resulting complex classification problem with hundreds thousands classes. Several methods have been proposed deal such problems, including several algorithms based on well-known bio-inspired techniques, as neural networks, ant colony optimization, genetic algorithms. In this work, we propose novel global method called GEHM, which makes use grammatical evolution for generating rules. approach, algorithm evolves antecedents rules, order assign instances dataset probabilistic class vector. Our compared protein function prediction datasets. The empirical analysis conducted work shows that GEHM outperforms statistical significance, suggests promising alternative multi-label biological data.