Abstract:In this work we put forward the inclusion of error mitigation routines in the process of training Variational Quantum Circuit (VQC) models. In detail, we define a Real Time Quantum Error Mitigation (RTQEM) algorithm to assist in fitting functions on quantum chips with VQCs. While state-of-the-art QEM methods cannot address the exponential loss concentration induced by noise in current devices, we demonstrate that our RTQEM routine can enhance VQCs' trainability by reducing the corruption of the loss function. We tested the algorithm by simulating and deploying the fit of a monodimensional $\textit{u}$-Quark Parton Distribution Function (PDF) on a superconducting single-qubit device, and we further analyzed the scalability of the proposed technique by simulating a multidimensional fit with up to 8 qubits.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the noise problem encountered when training variational quantum circuit (VQC) models on current quantum hardware. Specifically, the paper proposes a real - time quantum error mitigation (RTQEM) algorithm to improve the erosion of the loss function when using VQC for function fitting on noisy quantum chips. Traditional quantum error mitigation (QEM) methods cannot solve the problem of exponential loss concentration caused by noise, while RTQEM enhances the training ability of VQC by reducing the corruption of the loss function. ### Paper Background On current noisy intermediate - scale quantum (NISQ) devices, variational quantum algorithms (VQA) are one of the most promising methods in quantum machine learning (QML) models. These algorithms have been practically applied in fields such as quantum chemistry. However, hardware errors and long - running executions can disrupt the loss function landscape in the optimization process, leading to changes in the minimum position or the optimal loss value, thus limiting the practical applications of NISQ devices in certain algorithms. In addition, VQC models are also affected by noise - induced no - man's - land plateaus (NIBPs), resulting in vanishing gradients. ### Paper Goals To overcome these problems, the paper proposes the RTQEM algorithm, aiming to improve the training performance of VQC by mitigating noise in real - time. Specifically, the paper uses the important Clifford sampling (ICS) method, which is a learning - based quantum error mitigation technique. The core idea of ICS is to approximate the noise through parameterized mapping and continuously update this mapping during the training process to reduce the impact of noise on the gradient and the final prediction. ### Experimental Verification The paper verifies the effectiveness of the RTQEM algorithm through a series of one - dimensional and multi - dimensional regression problems. First, the author uses the VQC model to fit the one - dimensional u - quark parton distribution function (PDF), which is an important problem in high - energy physics. Then, the author further studies multi - dimensional fitting problems, simulating cases with up to 8 qubits to analyze the scalability of the algorithm. ### Main Contributions 1. **Propose the RTQEM algorithm**: By updating the noise model in real - time, reduce the impact of noise on the loss function, thereby improving the training performance of VQC. 2. **Experimental verification**: Through the simulation of one - dimensional and multi - dimensional regression problems and the deployment on actual superconducting qubits, the effectiveness of the RTQEM algorithm is verified. 3. **Analyze the impact of noise**: Analyze in detail the impact of noise on the optimization process and show how RTQEM improves the training results under different noise conditions. ### Conclusion The RTQEM algorithm proposed in the paper performs excellently in dealing with the noise problem on NISQ devices and can significantly improve the training performance of VQC models. This method provides strong support for future implementation of more complex quantum machine learning tasks on quantum hardware.

Real-time error mitigation for variational optimization on quantum hardware

Improving the speed of variational quantum algorithms for quantum error correction

Error mitigation in variational quantum eigensolvers using tailored probabilistic machine learning

Variational Quantum-Neural Hybrid Error Mitigation

Can Error Mitigation Improve Trainability of Noisy Variational Quantum Algorithms?

Bounding the systematic error in quantum error mitigation due to model violation

Variational Quantum-Neural Hybrid Eigensolver.

Noise-robust ground state energy estimates from deep quantum circuits

Synergetic quantum error mitigation by randomized compiling and zero-noise extrapolation for the variational quantum eigensolver

Volumetric Benchmarking of Error Mitigation with Qermit

Quantum subspace expansion in the presence of hardware noise

Quantum circuit architecture search: error mitigation and trainability enhancement for variational quantum solvers.

Comparative analysis of error mitigation techniques for variational quantum eigensolver implementations on IBM quantum system

Mitigating Quantum Gate Errors for Variational Eigensolvers Using Hardware-Inspired Zero-Noise Extrapolation

Error-Mitigated Digital Quantum Simulation.

Quantum error mitigation in quantum annealing

Adaptive quantum error mitigation using pulse-based inverse evolutions

Robust design under uncertainty in quantum error mitigation

Mitigating algorithmic errors in quantum optimization through energy extrapolation

Quantum error mitigation via quantum-noise-effect circuit groups