Robust Learning from Noisy Side-information by Semidefinite Programming
share
Robustness recently becomes one of the major concerns among machine learning community, since learning algorithms are usually vulnerable to outliers or corruptions. Motivated by such a trend and needs, we pursue robustness in semi-definite programming (SDP) in this paper. Specifically, this is done by replacing the commonly used squared loss with the more robust ℓ_1-loss in the low-rank SDP. However, the resulting objective becomes neither convex nor smooth. As no existing algorithms can be applied, we design an efficient algorithm, based on majorization-minimization, to optimize the objective. The proposed algorithm not only has cheap iterations and low space complexity but also theoretically converges to some critical points. Finally, empirical study shows that the new objective armed with proposed algorithm outperforms state-of-the-art in terms of both speed and accuracy.
READ FULL TEXT