-
Bounding the Graph Capacity with Quantum Mechanics and Finite Automata
Authors:
Alexander Meiburg
Abstract:
The zero-error capacity of a channel (or Shannon capacity of a graph) quantifies how much information can be transmitted with no risk of error. In contrast to the Shannon capacity of a channel, the zero-error capacity has not even been shown to be computable: we have no convergent upper bounds. In this work, we present a new quantity, the zero-error {\em unitary} capacity, and show that it can be…
▽ More
The zero-error capacity of a channel (or Shannon capacity of a graph) quantifies how much information can be transmitted with no risk of error. In contrast to the Shannon capacity of a channel, the zero-error capacity has not even been shown to be computable: we have no convergent upper bounds. In this work, we present a new quantity, the zero-error {\em unitary} capacity, and show that it can be succinctly represented as the tensor product value of a quantum game. By studying the structure of finite automata, we show that the unitary capacity is within a controllable factor of the zero-error capacity. This allows new upper bounds through the sum-of-squares hierarchy, which converges to the commuting operator value of the game. Under the conjecture that the commuting operator and tensor product value of this game are equal, this would yield an algorithm for computing the zero-error capacity.
△ Less
Submitted 16 March, 2024;
originally announced March 2024.
-
Generative Learning of Continuous Data by Tensor Networks
Authors:
Alex Meiburg,
**g Chen,
Jacob Miller,
Raphaëlle Tihon,
Guillaume Rabusseau,
Alejandro Perdomo-Ortiz
Abstract:
Beyond their origin in modeling many-body quantum systems, tensor networks have emerged as a promising class of models for solving machine learning problems, notably in unsupervised generative learning. While possessing many desirable features arising from their quantum-inspired nature, tensor network generative models have previously been largely restricted to binary or categorical data, limiting…
▽ More
Beyond their origin in modeling many-body quantum systems, tensor networks have emerged as a promising class of models for solving machine learning problems, notably in unsupervised generative learning. While possessing many desirable features arising from their quantum-inspired nature, tensor network generative models have previously been largely restricted to binary or categorical data, limiting their utility in real-world modeling problems. We overcome this by introducing a new family of tensor network generative models for continuous data, which are capable of learning from distributions containing continuous random variables. We develop our method in the setting of matrix product states, first deriving a universal expressivity theorem proving the ability of this model family to approximate any reasonably smooth probability density function with arbitrary precision. We then benchmark the performance of this model on several synthetic and real-world datasets, finding that the model learns and generalizes well on distributions of continuous and discrete variables. We develop methods for modeling different data domains, and introduce a trainable compression layer which is found to increase model performance given limited memory or computational resources. Overall, our methods give important theoretical and empirical evidence of the efficacy of quantum-inspired methods for the rapidly growing field of generative learning.
△ Less
Submitted 31 October, 2023;
originally announced October 2023.
-
Bounds on Irrationality Measures and the Flint-Hills Series
Authors:
Alex Meiburg
Abstract:
It is unknown whether the Flint-Hills series $\sum_{n=1}^\infty \frac{1}{n^3\sin^2(n)}$ converges. Alekseyev (2011) connected this question to the irrationality measure of $π$, that $μ(π) > \frac{5}{2}$ would imply divergence of the Flint-Hills series. In this paper we established a near-complete converse, that $μ(π) < \frac{5}{2}$ would imply convergence. The associated results on the density of…
▽ More
It is unknown whether the Flint-Hills series $\sum_{n=1}^\infty \frac{1}{n^3\sin^2(n)}$ converges. Alekseyev (2011) connected this question to the irrationality measure of $π$, that $μ(π) > \frac{5}{2}$ would imply divergence of the Flint-Hills series. In this paper we established a near-complete converse, that $μ(π) < \frac{5}{2}$ would imply convergence. The associated results on the density of close rational approximations may be of independent interest. The remaining edge case of $μ(π) = \frac{5}{2}$ is briefly addressed, with evidence that it would be hard to resolve.
△ Less
Submitted 28 August, 2022;
originally announced August 2022.
-
Reduction Rules and ILP Are All You Need: Minimal Directed Feedback Vertex Set
Authors:
Alex Meiburg
Abstract:
This note describes the development of an exact solver for Minimal Directed Feedback Vertex Set as part of the PACE 2022 competition. The solver is powered largely by aggressively trying to reduce the DFVS problem to a Minimal Cover problem, and applying reduction rules adapted from Vertex Cover literature. The resulting problem is solved as an Integer Linear Program (ILP) using SCIP. The resultin…
▽ More
This note describes the development of an exact solver for Minimal Directed Feedback Vertex Set as part of the PACE 2022 competition. The solver is powered largely by aggressively trying to reduce the DFVS problem to a Minimal Cover problem, and applying reduction rules adapted from Vertex Cover literature. The resulting problem is solved as an Integer Linear Program (ILP) using SCIP. The resulting solver performed the second-best in the competition, although a bug at submission time disqualified it. As an additional note, we describe a new vertex cover reduction generalizing the Desk reduction rule.
△ Less
Submitted 1 August, 2022;
originally announced August 2022.
-
Linear-time generalized Hartree-Fock algorithm for quasi-one-dimensional systems
Authors:
Alex Meiburg,
Bela Bauer
Abstract:
In many approximate approaches to fermionic quantum many-body systems, such as Hartree-Fock and density functional theory, solving a system of non-interacting fermions coupled to some effective potential is the computational bottleneck. In this paper, we demonstrate that this crucial computational step can be accelerated using recently developed methods for Gaussian fermionic matrix product states…
▽ More
In many approximate approaches to fermionic quantum many-body systems, such as Hartree-Fock and density functional theory, solving a system of non-interacting fermions coupled to some effective potential is the computational bottleneck. In this paper, we demonstrate that this crucial computational step can be accelerated using recently developed methods for Gaussian fermionic matrix product states (GFMPS). As an example, we study the generalized Hartree-Fock method, which unifies Hartree-Fock and self-consistent BCS theory, applied to Hubbard models with an inhomogeneous potential. We demonstrate that for quasi-one-dimensional systems with local interactions, our approach scales approximately linearly in the length of the system while yielding a similar accuracy to standard approaches that scale cubically in the system size.
△ Less
Submitted 17 December, 2021; v1 submitted 3 December, 2021;
originally announced December 2021.
-
Inapproximability of Positive Semidefinite Permanents and Quantum State Tomography
Authors:
Alex Meiburg
Abstract:
Matrix permanents are hard to compute or even estimate in general. It had been previously suggested that the permanents of Positive Semidefinite (PSD) matrices may have efficient approximations. By relating PSD permanents to a task in quantum state tomography, we show that PSD permanents are NP-hard to approximate within a constant factor, and so admit no FPTAS (unless P=NP). We also establish tha…
▽ More
Matrix permanents are hard to compute or even estimate in general. It had been previously suggested that the permanents of Positive Semidefinite (PSD) matrices may have efficient approximations. By relating PSD permanents to a task in quantum state tomography, we show that PSD permanents are NP-hard to approximate within a constant factor, and so admit no FPTAS (unless P=NP). We also establish that several natural tasks in quantum state tomography, even approximately, are NP-hard in the dimension of the Hilbert space. These state tomography tasks therefore remain hard even with only logarithmically few qubits.
△ Less
Submitted 5 April, 2022; v1 submitted 4 November, 2021;
originally announced November 2021.
-
Quantum Constraint Problems can be complete for $\mathsf{BQP}$, $\mathsf{QCMA}$, and more
Authors:
Alex Meiburg
Abstract:
A quantum constraint problem is a frustration-free Hamiltonian problem: given a collection of local operators, is there a state that is in the ground state of each operator simultaneously? It has previously been shown that these problems can be in P, NP-complete, MA-complete, or QMA_1-complete, but this list has not been shown to be exhaustive. We present three quantum constraint problems, that ar…
▽ More
A quantum constraint problem is a frustration-free Hamiltonian problem: given a collection of local operators, is there a state that is in the ground state of each operator simultaneously? It has previously been shown that these problems can be in P, NP-complete, MA-complete, or QMA_1-complete, but this list has not been shown to be exhaustive. We present three quantum constraint problems, that are (1) BQP_1-complete (also known as coRQP), (2) QCMA_1-complete and (3) coRP-complete. This provides the first natural complete problem for BQP_1. We also show that all quantum constraint problems can be realized on qubits, a trait not shared with classical constraint problems. These results suggest a significant diversity of complexity classes present in quantum constraint problems.
△ Less
Submitted 20 July, 2021; v1 submitted 20 January, 2021;
originally announced January 2021.
