Name
Abstract | ||
|---|---|---|
We present a multi-label multiple kernel learning (MKL) formulation in which the data are embedded into a low-dimensional space directed by the instance- label correlations encoded into a hypergraph. We formulate the problem in the kernel-induced feature space and propose to learn the kernel matrix as a linear combination of a given collection of kernel matrices in the MKL framework. The proposed learning formulation leads to a non-smooth min-max problem, which can be cast into a semi-infinite linear program (SILP). We further propose an ap- proximate formulation with a guaranteed error bound which involves an uncon- strained convex optimization problem. In addition, we show that the objective function of the approximate formulation is differentiable with Lipschitz continu- ous gradient, and hence existing methods can be employed to compute the optimal solution efficiently. We apply the proposed formulation to the automated annota- tion of Drosophila gene expression pattern images, and promising results have been reported in comparison with representative algorithms. |
Year | Venue | Keywords |
|---|---|---|
2008 | NIPS | convex optimization,objective function,feature space,lipschitz continuity,linear program |
Field | DocType | Citations |
Kernel (linear algebra),Mathematical optimization,Radial basis function kernel,Kernel embedding of distributions,Computer science,Multiple kernel learning,Polynomial kernel,Artificial intelligence,Kernel method,String kernel,Variable kernel density estimation,Machine learning | Conference | 32 |
PageRank | References | Authors |
1.80 | 12 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Shuiwang Ji | 1 | 2579 | 122.25 |
| Liang Sun | 2 | 500 | 24.61 |
| Rong Jin | 3 | 6206 | 334.26 |
| Jieping Ye | 4 | 6943 | 351.37 |