← All problems

Convergence conjecture for the coordinate-descent PARAFAC algorithm

9 years old · traced to

Let {Ak}\{\mathbf{A}^k\} be the sequence generated by Algorithm, where Ak\mathbf{A}^k is the solution (A(1),A(2),…,A(N))(\mathbf{A}^{(1)},\mathbf{A}^{(2)},\ldots,\mathbf{A}^{(N)}) in the kk-th iteration of the repeat loop. Assume that {Ak}\{\mathbf{A}^k\} is bounded. Convergence conjecture. The sequence {Ak}\{\mathbf{A}^k\} converges to a critical point Aˉ\bar{\mathbf{A}}. The convergence theorem for this multiconvex stochastic scheme is not clear; the conjecture asserts convergence under boundedness, while no proof or resolution is supplied here.

References
  1. Songting Shi, Xiang Li, Arkadiusz Sitek and Quanzheng Li, “Learning the Sparse and Low Rank PARAFAC Decomposition via the Elastic Net”, arXiv:1705.10015 (2017).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.