Evolving multi-label classification rules by exploiting high-order label correlations

[摘要] In multi-label classification tasks, each problem instance is associated with multiple classes simultaneously. In such settings, the correlation between labels contain valuable information that can be used to obtain more accurate classification models. The correlation between labels can be exploited in different levels such as capturing the pair-wise correlation or exploiting the higher-order correlations. Even though the high-order approach is more capable of modeling the correlation, it is computationally more demanding and has scalability issues. This paper aims at exploiting the high-order label correlations locally using supervised learning classifier systems (UCS). For this purpose, the label powerset (LP) strategy is employed and a prediction aggregation is utilized that improves the prediction capability of the LP method in the presence of unseen labelsets. Exact match ratio and Hamming loss measures are considered to evaluate the rule performance and the expected fitness value of individual classification rules is investigated using both metrics. Also, a computational complexity analysis is provided for training the proposed algorithm. The experimental results of the proposed method are compared with other well-known LP-based methods on multiple benchmark datasets and confirm the competitive performance of this method. (C) 2020 Elsevier B.V. All rights reserved.

[发布日期] 2020-12-05

[关键词] Multi-label classification;High-order label correlations;Label powerset;Learning classifier systems;Genetic algorithms