Discrete minimax estimation with trees

This work addresses estimation challenges in nonparametric statistics, but appears incremental as it builds on existing partitioning methods to achieve known optimal rates.

The authors tackled the problem of discrete minimax estimation by proposing a recursive data-based partitioning scheme for piecewise-constant or piecewise-linear density estimates, showing it achieves the optimal L1 minimax rate for certain discrete nonparametric classes.