Unsolved
Remove logarithmic gaps in minimax sample complexity for matrix normal covariance estimation
Sourced from the work of Rafael Mendes de Oliveira, William Cole Franks, Akshay Ramachandran, Michael Walter
§ Problem Statement
Setup
Unsolved Problem
§ Discussion
Loading discussion…
§ Significance & Implications
§ Known Partial Results
§ References
[1]
Near Optimal Sample Complexity for Matrix and Tensor Normal Models via Geodesic Convexity
Rafael Mendes de Oliveira, William Cole Franks, Akshay Ramachandran, Michael Walter (2021)
Annals of Statistics (to appear)
📍 Section 8 (Conclusion and open problems), unnumbered paragraph stating that for the matrix normal model the minimax bounds are only tight up to logarithmic factors and asking to close this gap (no numbered conjecture/remark; page not explicitly determined).
Source paper where this problem appears.