The spectral reconstruction conjecture for symmetric matrices

About 20 years old · traced to

Let AA be a real n×nn \times n symmetric matrix. For each ii, write AiA_i for the matrix obtained by deleting the iith row and column, and let eigen⁡(A)\operatorname{eigen}(A) denote the spectrum of AA. Spectral reconstruction conjecture. There exists a subgroup G(A)⊆O(n)G(A) \subseteq O(n) such that a real symmetric matrix BB satisfies

eigen⁡(B)=eigen⁡(A)\operatorname{eigen}(B)=\operatorname{eigen}(A)

and

eigen⁡(Bi)=eigen⁡(Ai)\operatorname{eigen}(B_i)=\operatorname{eigen}(A_i)

for each ii if and only if B=UAUtB=UAU^t for some U∈G(A)U \in G(A). The conjecture concerns whether the spectrum of a symmetric matrix together with the spectra of all principal vertex-deleted submatrices determines the matrix up to the prescribed orthogonal symmetries. The source notes that it is clearly true when rank⁡(A)=1\operatorname{rank}(A)=1, with G(A)G(A) choosable as the diagonal sign group Z2n\mathbb Z_2^n; in other cases the group may be a subgroup of the permutation group SnS_n.

References

Primary source

Hongyu He, “Eigenvectors and Reconstruction”, arXiv:math/0607142 (2006).

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.