Global Identifiability of Overcomplete Dictionary Learning via L1 and Volume Minimization

We propose a novel formulation for dictionary learning with an overcomplete dictionary, i.e., when the number of atoms is larger than the dimension of the dictionary. The proposed formulation consists of a weighted sum of ℓ 1 norms of the rows of the sparse coefficient matrix plus the log of the matrix volume of the dictionary matrix. The main contribution of this work is to show that this novel formulation guarantees global identifiability of the overcomplete dictionary, under a mild condition that the sparse coefficient matrix satisfies a strong scattering condition in the hypercube. Furthermore, if every column of the coefficient matrix is sparse and the dictionary guarantees ℓ 1 recovery, then the coefficient matrix is identifiable as well. This is a major breakthrough for not only dictionary learning but also general matrix factorization models as identifiability is guaranteed even when the latent dimension is higher than the ambient dimension. We also provide a probabilistic analysis and show that if the sparse coefficient matrix is generated from the widely adopted sparse-Gaussian model, then the 𝑚 × 𝑘 overcomplete dictionary is globally identifiable if the sample size is bigger than a constant times (𝑘 2 /𝑚) log(𝑘 2 /𝑚) with overwhelming probability. Finally, we propose an algorithm based on alternating minimization to solve the new proposed formulation.