Smale 论文列表: 多标签与标签分布学习方向

摘要: 本贴列出实验室在多标签和标签分布学习方向的一些工作. 主要目的是保存 bib 方便以后查阅与引用.

1. SCI 期刊论文

1.1 权重共享的多标记学习算法

• PR 至少算次顶刊. 这篇有四个审稿人.
• 同时应对稀疏标签、标签缺失、稀疏结构.
• 为特征矩阵和标签矩阵分别创建相关矩阵.
• 选择具有代表性的标签.
• 学习特征流行与标签流行.

@article{Qian2022Weight,
    author  = {Kun Qian and Xue-Yang Min and Yu-Sheng Cheng and Fan Min},
    title   = {Weight matrix sharing for multi-label learning},
    journal = {Pattern Recognition},
    volume  = {136}
    year    = {2023},
    pages   = {109156},
    doi     = {10.1016/j.patcog.2022.109156}
}

1.2 多标签主动学习算法 MASP

• 串并行神经网络: 串行部分进行特征提取, 并行部分进行每个标签的预测. 发表之后才发现, 它与多任务学习的基本方案一致.
• 该网络具有处理缺失标签的能力.
• 主动学习的时候根据密度和独立性选择代表样本.
• 一篇非常八股的论文, 从头到尾都是 1, 2, 3 这样的列举.

@article{Min2022Multi,
    author  = {Xue-Yang Min and Kun Qian and Ben-Wen Zhang and Guo-Jie Song and Fan Min},
    title   = {Multi-label active learning through serial--parallel neural networks},
    journal = {Knowledge-Based Systems},
    year    = {2022},
    volume  = {251},
    pages   = {109226},
    doi     = {10.1016/j.knosys.2022.109226}
}

1.3 具有双k标记恢复算法的自相关多标记学习算法

• 使用 $k$NN.
• 根据邻居标签, 进行标签填充.
• 使用特殊的概率矩阵计算及恢复.
• 产生新的 $k$ 个属性.
• 在新数据上学习线性分类器.
• 测试的时候根据邻居获得新的属性.

@article{Qian2022Self,
    author  = {Kun Qian and Xue-Yang Min and Yu-Sheng Cheng and Guo-Jie Song and Fan Min},
    title   = {Self-dependence multi-label learning with double k for missing labels},
    journal = {Artificial Intelligence Review},
    year    = {2022},
    pages   = {1--38},
    doi     = {10.1007/s10462-022-10279-1}
}

1.4 使用注意力机制来联合利用标签和实例信息

@article{Cheng2022Global,
    author  = {Yu-Sheng Cheng and Kun Qian and Fan Min},
    title   = {Global and local attention-based multi-label learning with missing labels},
    journal = {Information Sciences},
    year    = {2022},
    volume  = {594},
    pages   = {20--42},
    doi     = {10.1016/j.ins.2022.02.022}
}

1.5 基于余弦距离映射的多标签分布学习算法: COS-LDL

基于余弦距离映射来计算标签之间的相关性，并利用这种相关性来构建了一种新的标签分布学习方法.
引用次数 3.

@article{Zhang2020Cos,
    author  = {Heng-Ru Zhang and Yu-Ting Huang and Yuan-Yuan Xu and Fan Min},
    title   = {COS-LDL: Label distribution learning by cosine-based distance-mapping correlation},
    journal = {IEEE Access},
    year    = {2020},
    volume  = {8},
    pages   = {63961--63970},
    doi     = {10.1109/ACCESS.2020.2984622}
}

1.6 属性和标签分布驱动的多标签主动学习

@article{Wang2022Attribute,
    author  = {Min Wang and Tingting Feng and Zhaohui Shan and Fan Min},
    title   = {Attribute and label distribution driven multi-label active learning},
    journal = {Applied Intelligence},
    year    = {2022},
    volume  = {52},
    pages   = {11131--11146},
    doi     = {10.1007/s10489-021-03086-8}
}

1.7 基于随机游走的多标签符号值划分

@article{Wen2020Multi,
    author  = {Liu-Ying Wen and Chao-Guang Luo and Wei-Zhi Wu and Fan Min},
    title   = {Multi-label symbolic value partitioning through random walks},
    journal = {Neurocomputing},
    year    = {2020},
    volume  = {387},
    pages   = {195--209},
    doi     = {10.1016/j.neucom.2020.01.046}
}

2. 国际会议论文

2.1使用 GAN 进行数据扩增的标签分布学习算法

Label Distribution Learning with Data Augmentation using Generative Adversarial Networks

3. 核心期刊论文

3.1 融合标签局部相关性的标签分布学习算法

引用次数 0

@article{Rong2022Label,
    author  = {Bin-Yuan Rong and Yuan-Yuan Xu and Ya-Lan Lv and Heng-Ru Zhang},
    title   = {Label distribution learning by fusion of local correlation of labels},
    journal = {Journal of Shandong University (Natural Science)},
    year    = {2022},
    volume  = {57},
    number  = {7},
    pages   = {53--64}
    doi     = {10.6040/j.issn.1671-9352.4.2021.196}
}

3.2 基于三角距离相关性的标签分布学习算法

引用次数 0

@article{Huang2021Label,
    author  = {Yu-Ting Huang and Yuan-Yuan Xu and  Heng-Ru Zhang and Fan Min and others},
    title   = {Label distribution learning based on triangular distance correlation},
    journal = {CAAI Transactions on Intelligent System},
    year    = {2021},
    volume  = {16},
    number  = {3},
    pages   = {449--458}
    doi     = {10.11992/tis.202001027}
}

3.3 融合结构化标签依赖性的标签分布学习算法

引用次数 1

@article{Huang2020Label,
    author  = {Yu-Ting Huang and Yuan-Yuan Xu and  Heng-Ru Zhang and Fan Min},
    title   = {Label distribution learning by exploiting structural label dependency},
    journal = {Journal of Nanjing University Natural Science},
    year    = {2020},
    volume  = {56},
    number  = {4},
    pages   = {524--532},
    doi     = {10.13232/j.cnki.jnju.2020.04.010}
}

3.4 利用多特征混沌粒子滤波的视觉目标跟踪方法

引用次数 3

@article{Ma2020Visual,
    author  = {Yuan-Yuan Ma and Zheng-Yang Dang and Heng-Ru Zhang},
    title   = {Visual target tracking method using multi feature chaotic particle filter},
    journal = {Application Research of Computers},
    year    = {2020},
    volume  = {37},
    number  = {11},
    pages   = {3500--3503},
    doi     = {10.19734/j.issn.1001-3695.2019.06.0244}
}