Parameter-Free Spectral Kernel Learning
Qi Mao, Ivor Wai-Hung Tsang · arXiv (Cornell University) · 2012
Due to the growing ubiquity of unlabeled data, learning with unlabeled data is attracting increasing attention in machine learning. In this paper, we propose a novel semi-supervised kernel learning method which can seamlessly combine manifold structure of unlabeled data and Regularized Least-Squares (RLS) to learn a new kernel. Interestingly, the new kernel matrix can be obtained analytically with the use of spectral decomposition of graph Laplacian matrix. Hence, the proposed algorithm does not require any numerical optimization solvers. Moreover, by maximizing kernel target alignment on labeled data, we can also learn model parameters automatically with a closed-form solution. For a given graph Laplacian matrix, our proposed method does not need to tune any model parameter including the tradeoff parameter in RLS and the balance parameter for unlabeled data. Extensive experiments on ten benchmark datasets show that our proposed two-stage parameter-free spectral kernel learning algorithm can obtain comparable performance with fine-tuned manifold regularization methods in transductive setting, and outperform multiple kernel learning in supervised setting.
Cited by 8 works. doi:10.48550/arxiv.1203.3495 Open-access PDF
Authors
References
- Multiple kernels for object detection
- Semi-Supervised Learning
- Learning with Local and Global Consistency
- Multiple kernel learning, conic duality, and the SMO algorithm
- On Kernel Target Alignment
- Kernels and Regularization on Graphs
- Domain Transfer SVM for video concept detection
- Diffusion kernels on graphs and other discrete structures
- Hierarchical spatio-temporal context modeling for action recognition
- Learning with idealized kernels