Abstract:VQC can be understood through the lens of Fourier analysis. It is already well-known that the function space represented by any circuit architecture can be described through a truncated Fourier sum. We show that the spectrum available to that truncated Fourier sum is not entirely determined by the encoding gates of the circuit, since the variational part of the circuit can constrain certain coefficients to zero, effectively removing that frequency from the spectrum. To the best of our knowledge, we give the first description of the functional dependence of the Fourier coefficients on the variational parameters as trigonometric polynomials. This allows us to provide an algorithm which computes the exact spectrum of any given circuit and the corresponding Fourier coefficients. Finally, we demonstrate that by comparing the Fourier transform of the dataset to the available spectra, it is possible to predict which \gls{VQC} out of a given list of choices will be able to best fit the data.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to gain an in - depth understanding of the performance of variational quantum circuits (VQC) in supervised learning tasks through Fourier analysis and proposes a method to optimize the selection of VQC architectures suitable for specific datasets. Specifically, this paper mainly solves the following problems: 1. **Relationship between Spectrum and Coefficients**: - The paper reveals that the output of a variational quantum circuit can be represented as a finite Fourier sum of input parameters. Although it is known that the encoding gate determines the range of the spectrum, the variational part can constrain certain frequency coefficients to zero, thus affecting the actually available spectrum. - For the first time, the authors describe in detail the functional dependence between Fourier coefficients and variational parameters and propose an algorithm to calculate the exact spectrum of a given circuit and its corresponding Fourier coefficients. 2. **Selection of VQC Architectures**: - By comparing the Fourier transform of the dataset with the available spectra of different VQC architectures, the paper provides a method to predict which VQC is most suitable for fitting the data. This method can help researchers select the VQC architecture most suitable for a specific task instead of blindly increasing the circuit complexity. 3. **Trade - off between Expressivity and Trainability**: - The paper discusses the inherent trade - off between the expressivity and trainability of VQC. Although higher expressivity means stronger function approximation ability, it also makes it more difficult to find suitable parameters. Therefore, a method is needed to balance the two to ensure that VQC can represent data well and be effectively trained in practice. 4. **Experimental Verification**: - To verify the above theories, the paper designs a series of experiments to perform regression tasks on synthetic data and real - world data (such as the MNIST handwritten digit dataset) using different VQC architectures. Through these experiments, the researchers evaluate the performance of different architectures and rank them according to their spectral characteristics. ### Summary In general, this paper attempts to understand and optimize the application of variational quantum circuits in supervised learning through Fourier analysis, especially providing new insights and methods in how to select appropriate VQC architectures. This not only helps to improve the performance of VQC in practical tasks but also provides a theoretical basis for further research on quantum machine - learning algorithms. If you have more specific questions or need more detailed explanations, please feel free to let me know!

Fourier Analysis of Variational Quantum Circuits for Supervised Learning

Fourier expansion in variational quantum algorithms

Multidimensional Fourier series with quantum circuits

Synthesis of quantum circuits based on supervised learning and correlations

Learning capability of parametrized quantum circuits

Programming Variational Quantum Circuits with Quantum-Train Agent

Trainability and Expressivity of Hamming-Weight Preserving Quantum Circuits for Machine Learning

When Quantum and Classical Models Disagree: Learning Beyond Minimum Norm Least Square

Constrained and Vanishing Expressivity of Quantum Fourier Models

A Fourier analysis framework for approximate classical simulations of quantum circuits

Equivariant quantum circuits for learning on weighted graphs

A circuit-generated quantum subspace algorithm for the variational quantum eigensolver

Physics-Informed Bayesian Optimization of Variational Quantum Circuits

Learning Quantum Phase Estimation by Variational Quantum Circuits

Sub-universal variational circuits for combinatorial optimization problems

A semi-agnostic ansatz with variable structure for quantum machine learning

Fourier Series Guided Design of Quantum Convolutional Neural Networks for Enhanced Time Series Forecasting

Fourier-domain Variational Formulation and Its Well-posedness for Supervised Learning

Parametrized Quantum Circuits and their approximation capacities in the context of quantum machine learning

Supervised learning with quantum-enhanced feature spaces

Variational Quantum Approximate Spectral Clustering for Binary Clustering Problems