pith. sign in

arxiv: 1711.05174 · v1 · pith:MHKHTJ6Anew · submitted 2017-11-14 · 📊 stat.ML · cs.LG· stat.CO

Near-Optimal Discrete Optimization for Experimental Design: A Regret Minimization Approach

classification 📊 stat.ML cs.LGstat.CO
keywords designvarepsilonpointsalgorithmapproximationbestcriteriaefficiency
0
0 comments X
read the original abstract

The experimental design problem concerns the selection of k points from a potentially large design pool of p-dimensional vectors, so as to maximize the statistical efficiency regressed on the selected k design points. Statistical efficiency is measured by optimality criteria, including A(verage), D(eterminant), T(race), E(igen), V(ariance) and G-optimality. Except for the T-optimality, exact optimization is NP-hard. We propose a polynomial-time regret minimization framework to achieve a $(1+\varepsilon)$ approximation with only $O(p/\varepsilon^2)$ design points, for all the optimality criteria above. In contrast, to the best of our knowledge, before our work, no polynomial-time algorithm achieves $(1+\varepsilon)$ approximations for D/E/G-optimality, and the best poly-time algorithm achieving $(1+\varepsilon)$-approximation for A/V-optimality requires $k = \Omega(p^2/\varepsilon)$ design points.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Sequential Experimental Design for Transductive Linear Bandits

    stat.ML 2019-06 unverdicted novelty 7.0

    Introduces transductive linear bandits, gives instance-dependent lower bounds, and presents an algorithm matching them up to logarithmic factors, including the first non-asymptotic near-optimal method for standard lin...