'2014/02'에 해당되는 글 1

  1. 2014.02.13 Robust Matirx Factorization with Unknown Noise (ICCV, 2013)

Robust Matirx Factorization with Unknown Noise (ICCV, 2013)



  • Most popular loss functions include the L2 and L1 losses
  • L2 is optimal for Gaussian noise, while L1 is for Laplacian distributed noise
  • However, real data is often corrupted by an unknown noise distribution, which is unlikely to be purely Gaussian or Laplacian.
  • To address this problem, this paper proposes a low-rank matrix factorization problem with a Mixture of Gaussians (MoG) noise model.
  • The MoG model is a universal approximator for any continuous distribution, and hence is able to model a wider range of noise distributions.
  • The parameters of the proposed model, MoG, can be estimated with the traditional Expectation-Maximization (EM) under a Maximum Likelihood Estimation (MLE) framework.

MoG_ICCV_2013.pptx

 


신고
받은 트랙백이 없고 댓글이 없습니다.

티스토리 툴바