摘要: 通过近邻样例类标记确定测试样例类标记的思想在多标记分类算法中取得了良好的效果。该类算法通过对训练集进行学习,建立训练样例类标记与其k个近邻样例中不同类标记样例个数的映射关系,然后用该映射关系预测测试样例的类标记。该类算法的不足是只考虑近邻样例中不同类别样例的个数与测试样例类标记的映射关系,忽略了近邻样例与测试样例的局部相关性。考虑训练样例类与近邻样例的局部相关性,建立起它们类别间的映射关系,预测测试样例类标记,提出ML-WKNN算法。实验表明,ML-WKNN能更好地处理多标记分类问题和自动图像标注问题。
[1] McCallum Andrew.Multi-label text classification with a mix-ture model trained by EM[C]∥AAAI’99Workshop on Text Learning.1999:1-7 [2] Schapire,Robert E,Singer Y.BoosTexter:A boosting-basedsystem for text categorization[J].Machine learning,2000,39(2/3):135-168 [3] Tsoumakas,Grigorios,Katakis I.Multi-label classification:Anoverview[J].International Journal of Data Warehousing and Mining (IJDWM),2007,3(3):1-13 [4] Elisseeff,André,Weston J.A kernel method for multi-labelledclassification[J].Advances in neural information processing systems,2001,14:681-687 [5] Zhang Min-ling,Zhou Zhi-hua.ML-KNN:A lazy learning approach to multi-label learning[J].Pattern Recognition,2007,40(7):2038-2048 [6] 苗夺谦,卫志华.中文文本信息处理的原理与应用[M].北京:清华大学出版社,2007:219-228 [7] Clare,Amanda,Ding K R.Knowledge discovery in multi-labelphenotype data[C]∥Principles of Data Mining and Knowledge Discovery.2001:42-53 [8] Comité D,Francesco,Gilleron R,et al.Learning multi-label alternating decision trees from texts and data[C]∥Machine Learning and Data Mining in Pattern Recognition.2003:35-49 [9] Boutell,Matthew,Luo Jie-bo,et al.Learning multi-label scene classification[J].Pattern recognition,2004,7(9):1757-1771 [10] Zhang Min-ling,Zhou Zhi-hua.Multilabel neural networks with applications to functional genomics and text categorization[J].Knowledge and Data Engineering,2006, 18(10):1338-1351 [11] Mitchell T M.Machine Learning [M].USA:The McGrawHill Companies,Inc.1997:165-177 [12] 张学工.模式识别[M].北京:清华大学出版社,2010:120-130 [13] Boutell,Matthew,Luo Jie-bo,et al.Learning multi-label scene classification[J].Pattern recognition,2004,37(9):1757-1771 |
No related articles found! |
|