Abstract:Combinatorial reconfiguration is a growing research field studying problems on the transformability between a pair of solutions of a search problem. We consider the approximability of optimization variants of reconfiguration problems; e.g., for a Boolean formula $\varphi$ and two satisfying truth assignments $\sigma_{\sf s}$ and $\sigma_{\sf t}$ for $\varphi$, Maxmin SAT Reconfiguration requires to maximize the minimum fraction of satisfied clauses of $\varphi$ during transformation from $\sigma_{\sf s}$ to $\sigma_{\sf t}$. Solving such optimization variants approximately, we may obtain a reconfiguration sequence comprising almost-satisfying truth assignments. In this study, we prove a series of gap-preserving reductions to give evidence that a host of reconfiguration problems are PSPACE-hard to approximate, under some plausible assumption. Our starting point is a new working hypothesis called the Reconfiguration Inapproximability Hypothesis (RIH), which asserts that a gap version of Maxmin CSP Reconfiguration is PSPACE-hard. This hypothesis may be thought of as a reconfiguration analogue of the PCP theorem. Our main result is PSPACE-hardness of approximating Maxmin $3$-SAT Reconfiguration of bounded occurrence under RIH. The crux of its proof is a gap-preserving reduction from Maxmin Binary CSP Reconfiguration to itself of bounded degree. Because a simple application of the degree reduction technique using expander graphs due to Papadimitriou and Yannakakis does not preserve the perfect completeness, we modify the alphabet as if each vertex could take a pair of values simultaneously. To accomplish the soundness requirement, we further apply an explicit family of near-Ramanujan graphs and the expander mixing lemma. As an application of the main result, we demonstrate that under RIH, optimization variants of popular reconfiguration problems are PSPACE-hard to approximate.

Scenario reduction revisited: fundamental limits and guarantees

Scenario Reduction Algorithms in Stochastic Programming

Scenario Reduction Network Based on Wasserstein Distance with Regularization

Scenario Tree Reduction via Wasserstein Barycenters

Scenario Reduction with Guarantees for Stochastic Optimal Control of Linear Systems

Problem-Driven Scenario Reduction and Scenario Approximation for Robust Optimization

Scenario approach for minmax optimization with emphasis on the nonconvex case: positive results and caveats

Efficient Algorithms for Distributionally Robust Stochastic Optimization with Discrete Scenario Support

Quantitative reductions and vertex-ranked infinite games

Scenario Reduction Based on Correlation Sensitivity and Its Application in Microgrid Optimization

Gap Preserving Reductions Between Reconfiguration Problems

Quasi-stationary distributions in reducible state spaces

Efficient Scenario Generation for Heavy-tailed Chance Constrained Optimization

Automatic scenario generation for efficient solution of robust optimal control problems

Sample Complexity of Chance Constrained Optimization in Dynamic Environment

Margin theory for the scenario-based approach to robust optimization in high dimension

Computationally efficient reductions between some statistical models

A Note on Gale, Kuhn, and Tucker's Reductions of Zero-Sum Games

Discretely Beyond $1/e$: Guided Combinatorial Algorithms for Submodular Maximization

Finer-grained Reductions in Fine-grained Hardness of Approximation

Minimising the Probabilistic Bisimilarity Distance