Overcomplete Dictionary Learning for Nonnegative Sparse Representation with an ?_p-Norm Constraint Based on Majorize-Minimization

摘要

This paper is for addressing the nonnegative sparse representation problem, i.e. to represent a nonnegative matrix as an over complete nonnegative dictionary times a nonnegative coefficient matrix. Many data in the real world can be represented sparsely by combinations of typical features. Moreover, large family of nonnegative data, such as image pixels, word frequency, power spectrum etc., are in great demand for engineering problems. Nonnegative Sparse Representation (NSR) is attractive to nonnegative data analysis. In this study, Overcomplete dictionary learning for NSR with an ?_p-norm (0 <; p <;1) is proposed and the ?_p-norm is expected for leading a sparser solution than ?₁-norm. An ?_p-norm is non-convex, then ?_p-norm was approximated by weighted ?₁-norm for convex optimization. We performed experiments about recovering accuracies of dictionary and coefficients. The recovering ratio is evaluated for various sparse levels of coefficients. The average of recovery ratios by proposed method for each sparse level were higher compared with an unweighted ?₁-norm. It improved +15.37% at most. We have shown that the proposed method also has advantages in recovering sparseness, ?₂ relative error and support distance of coefficients etc.