-
Quantum Adversarial Learning in Emulation of Monte-Carlo Methods for Max-cut Approximation: QAOA is not optimal
Authors:
Cem M. Unsal,
Lucas T. Brady
Abstract:
One of the leading candidates for near-term quantum advantage is the class of Variational Quantum Algorithms, but these algorithms suffer from classical difficulty in optimizing the variational parameters as the number of parameters increases. Therefore, it is important to understand the expressibility and power of various ansätze to produce target states and distributions. To this end, we apply n…
▽ More
One of the leading candidates for near-term quantum advantage is the class of Variational Quantum Algorithms, but these algorithms suffer from classical difficulty in optimizing the variational parameters as the number of parameters increases. Therefore, it is important to understand the expressibility and power of various ansätze to produce target states and distributions. To this end, we apply notions of emulation to Variational Quantum Annealing and the Quantum Approximate Optimization Algorithm (QAOA) to show that QAOA is outperformed by variational annealing schedules with equivalent numbers of parameters. Our Variational Quantum Annealing schedule is based on a novel polynomial parameterization that can be optimized in a similar gradient-free way as QAOA, using the same physical ingredients. In order to compare the performance of ansätze types, we have developed statistical notions of Monte-Carlo methods. Monte-Carlo methods are computer programs that generate random variables that approximate a target number that is computationally hard to calculate exactly. While the most well-known Monte-Carlo method is Monte-Carlo integration (e.g. Diffusion Monte-Carlo or path-integral quantum Monte-Carlo), QAOA is itself a Monte-Carlo method that finds good solutions to NP-complete problems such as Max-cut. We apply these statistical Monte-Carlo notions to further elucidate the theoretical framework around these quantum algorithms.
△ Less
Submitted 24 November, 2022;
originally announced November 2022.
-
Faster Quantum Concentration via Grover's Search
Authors:
Cem M. Unsal,
A. Yavuz Oruc
Abstract:
We present quantum algorithms for routing concentration assignments on full capacity fat-and-slim concentrators, bounded fat-and-slim concentrators, and regular fat-and-slim concentrators. Classically, the concentration assignment takes $O(n)$ time on all these concentrators, where $n$ is the number of inputs. Powered by Grover's quantum search algorithm, our algorithms take $O(\sqrt{nc}\ln{c})$ t…
▽ More
We present quantum algorithms for routing concentration assignments on full capacity fat-and-slim concentrators, bounded fat-and-slim concentrators, and regular fat-and-slim concentrators. Classically, the concentration assignment takes $O(n)$ time on all these concentrators, where $n$ is the number of inputs. Powered by Grover's quantum search algorithm, our algorithms take $O(\sqrt{nc}\ln{c})$ time, where $c$ is the capacity of the concentrator. Thus, our quantum algorithms are asymptotically faster than their classical counterparts, when $c\ln^2{c}=o(n)$.In general, $c = n^μ,$ satisfies $c\ln^2{c}=o(n),$ implying a time complexity of $O(n^{0.5(1+ μ)} \ln n),$ for any $μ, 0 < μ< 1.$
△ Less
Submitted 16 March, 2021;
originally announced March 2021.
-
Circuit Transformations for Quantum Architectures
Authors:
Andrew M. Childs,
Eddie Schoute,
Cem M. Unsal
Abstract:
Quantum computer architectures impose restrictions on qubit interactions. We propose efficient circuit transformations that modify a given quantum circuit to fit an architecture, allowing for any initial and final mapping of circuit qubits to architecture qubits. To achieve this, we first consider the qubit movement subproblem and use the routing via matchings framework to prove tighter bounds on…
▽ More
Quantum computer architectures impose restrictions on qubit interactions. We propose efficient circuit transformations that modify a given quantum circuit to fit an architecture, allowing for any initial and final mapping of circuit qubits to architecture qubits. To achieve this, we first consider the qubit movement subproblem and use the routing via matchings framework to prove tighter bounds on parallel routing. In practice, we only need to perform partial permutations, so we generalize routing via matchings to that setting. We give new routing procedures for common architecture graphs and for the generalized hierarchical product of graphs, which produces subgraphs of the Cartesian product. Secondly, for serial routing, we consider the token swapping framework and extend a 4-approximation algorithm for general graphs to support partial permutations. We apply these routing procedures to give several circuit transformations, using various heuristic qubit placement subroutines. We implement these transformations in software and compare their performance for large quantum circuits on grid and modular architectures, identifying strategies that work well in practice.
△ Less
Submitted 6 September, 2019; v1 submitted 25 February, 2019;
originally announced February 2019.
