A novel algorithm based on optimal approximate design theory selects nearly optimal subdata for parametric model parameter estimation, with a convergence proof and tight efficiency bounds that outperform prior methods.
The information matrix ofξP can be viewed as a function ofwww′=(w1,...,wn2−1), wherewi denotes the weight forxi∈X2, i = 1,...,n 2−1
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Nearly Optimal Subdata Selection
A novel algorithm based on optimal approximate design theory selects nearly optimal subdata for parametric model parameter estimation, with a convergence proof and tight efficiency bounds that outperform prior methods.