다중레이블 학습을 위한 희소성과 저랭크 기반 특징 선별 기법
Sparse and low-rank feature selection for multi-label learning
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 논문에서는 다중 레이블 분류를 위한 특징 선별 기법을 제안한다. 기존 많은 특징 선별 기법들은상호정보척도 등을 이용하여 특징과 레이블 사이의 연관성을 계산하여 특징을 선별하였다. 하지만상호정보척도는 결합 확률을 요구하기 때문에 실제 전제 특징 집합에서 결합 확률을 계산하는 것은어렵다. 따라서 소수의 특징만 계산이 가능하여 지역적 최적화만 가능하다는 단점을 가진다. 이런 지역적 최적화 문제를 피해, 주어진 특징 전체 공간에서 저랭크 공간을 구성하고, 희소성을 가진 특징들을선별할 수 있는 특징 선별 기법을 제안한다. 이를 위해 뉴클리어 노름을 이용해 회귀 기반의 목적함수를 설계하였고, 이 목적 함수의 최적화 문제를 풀기 위한 경사하강법 방식의 알고리즘을 제안하였다. 4가지의 데이터와 3가지 다중 레이블 분류 성능을 기준으로 다중 레이블 분류 실험 결과를 통해제안하는 방법론이 기존 특징 선별 기법보다 좋은 성능을 나타내는 것을 보였다. 또한 제안하는 목적함수의 파라미터 값 변화에도 성능 변화가 둔감한 것을 실험적인 결과로 확인하였다.
Abstract
In this paper, we propose a feature selection technique for multi-label classification. Many existing feature selection techniques have selected features by calculating the relation between features and labels such as a mutual information scale. However, since the mutual information measure requires a joint probability, it is difficult to calculate the joint probability from an actual premise feature set. Therefore, it has the disadvantage that only a few features can be calculated and only local optimization is possible. Away from this regional optimization problem, we propose a feature selection technique that constructs a low-rank space in the entire given feature space and selects features with sparsity. To this end, we designed a regression-based objective function using Nuclear norm, and proposed an algorithm of gradient descent method to solve the optimization problem of this objective function. Based on the results of multi-label classification experiments on four data and three multi-label classification performance, the proposed methodology showed better performance than the existing feature selection technique. In addition, it was showed by experimental results that the performance change is insensitive even to the parameter value change of the proposed objective function.
참고문헌 (23)
- Ronghua Shang, "Local discriminative based sparse subspace learning for feature selection", Pattern Recognition, 92, pp.219~230 (2019), Elsevier BV[DOI]
- Xiaofeng Zhu, "Low-Rank Sparse Subspace for Spectral Clustering", IEEE Transactions on Knowledge and Data Engineering, 31(8), pp.1532~1543 (2019), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Yong Zhang, "Nonnegative Laplacian embedding guided subspace learning for unsupervised feature selection", Pattern Recognition, 93, pp.337~352 (2019), Elsevier BV[DOI]
- Sadia Sharmin, "Simultaneous feature selection and discretization based on mutual information", Pattern Recognition, 91, pp.162~174 (2019), Elsevier BV[DOI]
- Rafael B. Pereira, "Categorizing feature selection methods for multi-label classification", Artificial Intelligence Review, 49(1), pp.57~78 (2018), Springer Science and Business Media LLC[DOI]
- Yaojin Lin, "Multi-label feature selection based on max-dependency and min-redundancy", Neurocomputing, 168, pp.92~103 (2015), Elsevier BV[DOI]
- J. Lee, "Approximating mutual information for multi-label feature selection", Electronics Letters, 48(15), pp.929~ (2012), Institution of Engineering and Technology (IET)[DOI]
- Yaojin Lin, "Multi-label feature selection based on max-dependency and min-redundancy", Neurocomputing, 168, pp.92~103 (2015), Elsevier BV[DOI]
- S. Jungjit, "Two extensions to multi-lable correlation-based feature selection : A case study in bioinformatics", IEEE Internaional Conference on Systems, Man, and Cybernetics, pp.1519~1524 (2013)
- Hyunki Lim, "Optimization approach for feature selection in multi-label classification", Pattern Recognition Letters, 89, pp.25~30 (2017), Elsevier BV[DOI]
- Zheng Zhang, "A Survey of Sparse Representation: Algorithms and Applications", IEEE Access, 3, pp.490~530 (2015), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Hengmin Zhang, "Low-Rank Matrix Recovery via Modified Schatten-$p$ Norm Minimization With Convergence Guarantees", IEEE Transactions on Image Processing, 29, pp.3132~3142 (2020), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Xiantong Zhen, "Multi-Target Regression via Robust Low-Rank Learning", IEEE Transactions on Pattern Analysis and Machine Intelligence, 40(2), pp.497~504 (2018), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- D. P. Bertsekas, "Constrained optimization and Lagrange multiplier methods" (2014), Academic press
- Min-Ling Zhang, "ML-KNN: A lazy learning approach to multi-label learning", Pattern Recognition, 40(7), pp.2038~2048 (2007), Elsevier BV[DOI]
- Min-Ling Zhang, "Feature selection for multi-label naive Bayes classification", Information Sciences, 179(19), pp.3218~3229 (2009), Elsevier BV[DOI]
- Min-Ling Zhang, "A Review on Multi-Label Learning Algorithms", IEEE Transactions on Knowledge and Data Engineering, 26(8), pp.1819~1837 (2014), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- K. Trohidis, "Multi-label classification of music into emotions", ISMIR, pp.325~330 (2008)
- F. Briggs, "New methods for acoustic classification of multiple simutaneous bird species in a noisy environment", IEEE International Workshop on Machine Learning for Signal Processing, pp.1~8 (2013)
- Matthew R. Boutell, "Learning multi-label scene classification", Pattern Recognition, 37(9), pp.1757~1771 (2004), Elsevier BV[DOI]
- A. Elisseeff, "A kernel method for multi-labelled classification", NIPS, 681(687), pp.681~687 (2001)
- Alberto Cano, "LAIM discretization for multi-label data", Information Sciences, 330, pp.370~384 (2016), Elsevier BV[DOI]
- Jaesung Lee, "SCLS: Multi-label feature selection based on scalable criterion for large label set", Pattern Recognition, 66, pp.342~352 (2017), Elsevier BV[DOI]