-
Independence and Induction in Reverse Mathematics
Authors:
David Belanger,
Chi Tat Chong,
Rupert Hölzl,
Frank Stephan
Abstract:
We continue the project of the study of reverse mathematics principles inspired by cardinal invariants. In this article in particular we focus on principles encapsulating the existence of large families of objects that are in some sense mutually independent. More precisely, we study the principle $\mathsf{MAD}$ stating that a maximal family of pairwise almost disjoint sets exists; and the principl…
▽ More
We continue the project of the study of reverse mathematics principles inspired by cardinal invariants. In this article in particular we focus on principles encapsulating the existence of large families of objects that are in some sense mutually independent. More precisely, we study the principle $\mathsf{MAD}$ stating that a maximal family of pairwise almost disjoint sets exists; and the principle $\mathsf{MED}$ expressing the existence of a maximal family of functions that are pairwise eventually different. We investigate characterisations of and relations between these principles and some of their variants. It turns out that induction strength at the levels of $\mathsf{B}\mathrmΣ_2^0$ or $\mathsf{I}\mathrmΣ_2^0$ is an essential parameter; for instance, over $\mathsf{B}\mathrmΣ_2^0$, we show that $\neg\mathsf{MAD}$ is equivalent to the principle $\mathsf{DOM}$ expressing that every weakly represented family of functions is dominated by some other function.
△ Less
Submitted 7 February, 2025; v1 submitted 19 August, 2024;
originally announced August 2024.
-
Conservation theorems for the Cohesiveness Principle
Authors:
David R. Belanger
Abstract:
We prove that the Cohesiveness Principle (COH) is $Π^1_1$ conservative over $RCA_0 + IΣ^0_n$ and over $RCA_0 + BΣ^0_n$ for all $n \geq 2$ by recursion-theoretic means. We first characterize COH over $RCA_0 + BΣ^0_2$ as a `jumped' version of Weak König's Lemma (WKL) and develop suitable machinery including a version of the Friedberg jump-inversion theorem. The main theorem is obtained when we combi…
▽ More
We prove that the Cohesiveness Principle (COH) is $Π^1_1$ conservative over $RCA_0 + IΣ^0_n$ and over $RCA_0 + BΣ^0_n$ for all $n \geq 2$ by recursion-theoretic means. We first characterize COH over $RCA_0 + BΣ^0_2$ as a `jumped' version of Weak König's Lemma (WKL) and develop suitable machinery including a version of the Friedberg jump-inversion theorem. The main theorem is obtained when we combine these with known results about WKL. In an appendix we give a proof of the $Π^1_1$ conservativity of WKL over $RCA_0$ by way of the Superlow Basis Theorem and a new proof of a recent jump-inversion theorem of Towsner.
△ Less
Submitted 25 December, 2022;
originally announced December 2022.
-
Constrained Discrete Black-Box Optimization using Mixed-Integer Programming
Authors:
Theodore Papalexopoulos,
Christian Tjandraatmadja,
Ross Anderson,
Juan Pablo Vielma,
David Belanger
Abstract:
Discrete black-box optimization problems are challenging for model-based optimization (MBO) algorithms, such as Bayesian optimization, due to the size of the search space and the need to satisfy combinatorial constraints. In particular, these methods require repeatedly solving a complex discrete global optimization problem in the inner loop, where popular heuristic inner-loop solvers introduce app…
▽ More
Discrete black-box optimization problems are challenging for model-based optimization (MBO) algorithms, such as Bayesian optimization, due to the size of the search space and the need to satisfy combinatorial constraints. In particular, these methods require repeatedly solving a complex discrete global optimization problem in the inner loop, where popular heuristic inner-loop solvers introduce approximations and are difficult to adapt to combinatorial constraints. In response, we propose NN+MILP, a general discrete MBO framework using piecewise-linear neural networks as surrogate models and mixed-integer linear programming (MILP) to optimize the acquisition function. MILP provides optimality guarantees and a versatile declarative language for domain-specific constraints. We test our approach on a range of unconstrained and constrained problems, including DNA binding, constrained binary quadratic problems from the MINLPLib benchmark, and the NAS-Bench-101 neural architecture search benchmark. NN+MILP surpasses or matches the performance of black-box algorithms tailored to the constraints at hand, with global optimization of the acquisition problem running in a few minutes using only standard software packages and hardware.
△ Less
Submitted 14 June, 2022; v1 submitted 18 October, 2021;
originally announced October 2021.
-
Where Pigeonhole Principles meet König Lemmas
Authors:
David Belanger,
Chitat Chong,
Wei Wang,
Tin Lok Wong,
Yue Yang
Abstract:
We study the pigeonhole principle for $Σ_2$-definable injections with domain twice as large as the codomain, and the weak König lemma for $Δ^0_2$-definable trees in which every level has at least half of the possible nodes. We show that the latter implies the existence of $2$-random reals, and is conservative over the former. We also show that the former is strictly weaker than the usual pigeonhol…
▽ More
We study the pigeonhole principle for $Σ_2$-definable injections with domain twice as large as the codomain, and the weak König lemma for $Δ^0_2$-definable trees in which every level has at least half of the possible nodes. We show that the latter implies the existence of $2$-random reals, and is conservative over the former. We also show that the former is strictly weaker than the usual pigeonhole principle for $Σ_2$-definable injections.
△ Less
Submitted 7 December, 2019;
originally announced December 2019.
