Abstract:Multi-label classification handles scenarios where an instance can be annotated with multiple non-exclusive but semantically related labels simultaneously. Despite significant progress, multi-label classification is still challenging due to the emergence of multiple applications leading to high-dimensional label spaces. Researchers have generalized feature dimensionality reduction techniques to label space by using label correlation information, and obtained two techniques: label embedding and label selection. There have been many successful algorithms in label embedding, but less attention has been paid to label selection. In this paper, we propose a label selection algorithm for multi-label classification: LS-AntRL, which combines ant colony optimization (ACO) and reinforcement learning (RL). This method helps ant colony algorithms search better in the search space by using temporal difference (TD) RL algorithm to learn directly from the experience of ants. For heuristic learning, we need to model the ACO problem as a RL problem, that is, to model label selection as a Markov decision process (MDP), where the label represents the state, and each ant selecting unvisited labels represents a set of actions. The state transition rules of the ACO algorithm constitute the transition function in the MDP, and the state value function is updated by TD formula to form a heuristic function in ACO. After performing label selection, we train a binary weighted neural network to recover low-dimensional label space back to the original label space. We apply the above model to five benchmark datasets with more than 100 labels. Experimental results show that our method achieves better classification performance than other advanced methods in terms of two performance evaluation metrics (Precision@n and DCG@n).

Multi-Label Classification Algorithm Using Adaptive Linear Regression

A Label-Specific Attention-Based Network with Regularized Loss for Multi-label Classification

Adaptive Low-Rank Multi-Label Active Learning for Image Classification.

Adaptive label thresholding methods for online multi-label classification.

Label-Related Adaptive Graph Construction Based on Attention for Multi-label Text Classification

Multi-label classification by exploiting label correlations

A survey of multi-label classification based on supervised and semi-supervised learning

An Empirical Comparison of Label Detection Techniques for Multi-Label Classification

Label Selection Algorithm Based on Ant Colony Optimization and Reinforcement Learning for Multi-label Classification.

Research of multi-label text classification based on label attention and correlation networks

A Multi-Label Learning Algorithm Based on Sparse Representation

A Multi Label Learning Method Using Affinity Propagation and Support Vector Machine

MsCoa: Multi-Step Co-Attention Model for Multi-Label Classification

Multi-label learning by LLE dimension reduction and Bayesian classification

Long-tailed Multi-label Text Classification Via Label Co-occurrence-Aware Knowledge Transfer

A Multi-Label Classification Algorithm Based on Label Set Partition

Joint Multi-Label Classification and Label Correlations with Missing Labels and Feature Selection

Multi-Label Classification Algorithm Based on Association Rule Mining

Adaptively Weighted Copy-Decoupling Resampling Strategy for Long-Tailed Multi-label Classification

Multi-label Text Classification Based on the Label Correlation Mixture Model.

Multi-label classification with weak labels by learning label correlation and label regularization