Rethinking the symmetry-preserving circuits for constrained variational quantum algorithms
Authors: Ge Yan, Hongxu Chen, Kaisen Pan, Junchi Yan
ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conduct detailed numerical experiments on two well-studied symmetry-preserving problems, namely ground state energy estimation and feature selection in machine learning. The superior performance demonstrates the efficiency and supremacy of the proposed HW preserving ansatz on constrained VQAs. |
| Researcher Affiliation | Academia | Ge Yan, Hongxu Chen, Kaiseng Pan, Junchi Yan Department of Computer Science and Engineering, Shanghai Jiao Tong University {yange98,mechta chx,pks0813,yanjunchi}@sjtu.edu.cn |
| Pseudocode | Yes | The detailed computational method to calculate the dimension of DLA is described in Alg. 1 (Schirmer et al., 2001). Algorithm 1 Computing the dimension of DLA |
| Open Source Code | No | The paper does not include an explicit statement or a link indicating the release of source code for the described methodology. |
| Open Datasets | Yes | We obtain the Hamiltonian of these molecules from the Python package Open Fermion (Mc Clean et al., 2020). We select four open-source datasets to test the HW preserving ansatz as well as a classical QUBO solver py QUBO (Tanahashi et al., 2019; Zaman et al., 2021). Statistics for the datasets are listed in Tab. 2. Wine Quality (Cortez et al., 2009), Heart Disease (Janosi et al., 1988), Titanic (Vanschoren et al., 2013), Dry Bean (Koklu & Ozkan, 2020). |
| Dataset Splits | No | The paper does not provide specific train/validation/test dataset splits (e.g., percentages, sample counts, or references to predefined splits). |
| Hardware Specification | Yes | Experiments are performed on a machine with 190GB memory, one physical CPU with 32 cores AMD Ryzen 3970X CPU, 5 GPUs (NV Ge Force RTX 3090). |
| Software Dependencies | No | The paper mentions using 'Open Fermion', 'py QUBO', and 'qiskit.quantum info.random unitary in IBM Qiskit' but does not provide specific version numbers for these software components. |
| Experiment Setup | Yes | We first use a fixed bond length and vary the number of layers. We fix the number of parameters for each method to the minimum value that reaches overparameterization. We set k for all the datasets as 3 indicating that we select the top 3 features from all the datasets. α is the penalty coefficient. |