Non-asymptotic Approximation Error Bounds of Parameterized Quantum Circuits

Authors: Zhan Yu, Qiuhao Chen, Yuling Jiao, Yinan Li, Xiliang Lu, Xin Wang, Jerry Yang

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We further validate the approximation capability of PQCs through numerical experiments.
Researcher Affiliation Academia 1 School of Mathematics and Statistics, Wuhan University, Wuhan 430072, China 2 Centre for Quantum Technologies, National University of Singapore, 117543, Singapore 3 Hubei Key Laboratory of Computational Science, Wuhan 430072, China 4 Thrust of Artificial Intelligence, Information Hub, Hong Kong University of Science and Technology (Guangzhou), Guangzhou 511453, China
Pseudocode No The paper includes circuit diagrams (e.g., Figure 1) and detailed descriptions of algorithms and constructions in prose and mathematical notation but does not contain explicitly labeled "Pseudocode" or "Algorithm" blocks.
Open Source Code Yes We have provided the complete code in the supplementary materials that is necessary to reproduce the experimental results.
Open Datasets No We randomly sample 200 data points within the domain [0, 1] to create training and test datasets for D(x).
Dataset Splits No The paper mentions 'training and test datasets' but does not specify a separate validation split or explicit percentages/counts for training, validation, and test datasets.
Hardware Specification Yes Both learning processes are implemented on a Gold 6248 2.50 GHz Intel(R) Xeon(R) CPU.
Software Dependencies No The paper mentions using the 'Adam optimizer' but does not provide specific version numbers for software components or libraries.
Experiment Setup Yes Each parameter of the PQC is randomly initialized within the range [0, π]. We use the Adam optimizer [55] with a learning rate of 0.01 to minimize the Mean Squared Error (MSE) loss function during training. The training process was limited to a maximum of 300 iterations with a batch size of 100 data points. Early termination occurred if the MSE reached below 10−4.