Title: Quantum-Enhanced Stochastic Gradient Descent for Deep Learning

Jincheng Zhang

Published Sep 7, 2026• 0 citations

Abstract

Quantum-enhanced stochastic gradient descent (SGD) presents a promising avenue for accelerating deep learning training and enhancing generalization, leveraging quantum phenomena such as superposition and entanglement to potentially improve convergence rates and model robustness. This paper explores the feasibility of utilizing quantum states to represent and update neural network weights, proposing a novel algorithm that aims to address the limitations of current SGD approaches in handling large datasets and intricate models. We investigate the potential benefits of quantum acceleration in deep learning, focusing on the enhancement of learning speed and improved model performance. The paper details the core mechanism, outlines the experimental setup, and discusses the challenges and future directions of this research.

DOI: 10.5281/zenodo.22579511

Related papers

Title: Quantum-Inspired Optimization via Stochastic Gradient Descent

Jincheng Zhang

Quantum-inspired optimization represents a promising avenue for addressing computationally challenging optimization problems. This paper investigates the application of stochastic gradient descent (SGD) to emulate the quantum annealing process, aiming to improve the speed and accuracy of complex optimization tasks. We propose a novel algorithm that iteratively refines a solution based on the gradient of the objective function, mimicking the probabilistic exploration inherent in quantum annealing. This approach offers a potentially more efficient and scalable method compared to traditional optimization techniques, leveraging the strengths of stochasticity. The core mechanism centers around utilizing SGD to dynamically adjust the solution trajectory, effectively mimicking the quantum annealing process.

Aug 31, 20260 citations

Quantum-Inspired Stochastic Gradient Descent for Deep Learning

Jincheng Zhang

Stochastic Gradient Descent (SGD) is a cornerstone algorithm in deep learning, widely employed for optimizing model parameters. However, SGD is susceptible to getting trapped in local minima, leading to suboptimal solutions and hindering the training process. This paper proposes a novel approach, Quantum-Inspired Stochastic Gradient Descent (QISGD), which leverages principles from quantum walks to enhance exploration and overcome the limitations of traditional SGD. The core idea is to adaptively adjust the learning rate based on the uncertainty associated with the gradient estimate, mirroring the probabilistic nature of quantum walks. Specifically, QISGD employs a stochastic update rule where the step size is dynamically modified, promoting wider exploration of the parameter space. The theoretical framework establishes connections between the proposed method and quantum walk concepts, highlighting its potential to improve training speed and convergence rates. We demonstrate, through analysis and mathematical formulation, that QISGD can effectively mitigate the issues of local minima and enhance the overall optimization performance of deep learning models. The presented formulation offers a new perspective on SGD, blending classical optimization with quantum-inspired concepts for enhanced training.

Aug 31, 20260 citations

Quantum-Inspired Stochastic Gradient Descent for Large-Scale Neural Networks

Jincheng Zhang

Training large-scale neural networks with traditional stochastic gradient descent (SGD) methods often faces significant challenges, including slow convergence rates and high computational costs. This paper proposes a novel approach, Quantum-Inspired Stochastic Gradient Descent (QISGD), designed to address these limitations. QISGD leverages principles inspired by quantum measurement processes, specifically employing a random projection operator to approximate the gradient. This technique aims to provide a more efficient and potentially faster way to update the network's parameters during training. The theoretical underpinnings of QISGD are presented, along with a detailed explanation of the algorithm's components. We demonstrate the potential benefits of QISGD through a conceptual analysis and outline a framework for future research and implementation. The core innovation lies in translating quantum measurement intuition into a practical optimization strategy for classical neural networks. The goal is to improve the training efficiency of large networks while retaining the benefits of SGD.

Aug 28, 20260 citations

Title: Quantum-Inspired Non-Linearity in Machine Learning

Jincheng Zhang

Quantum-inspired non-linearities offer a promising avenue for enhancing the robustness and generalization of machine learning models. This paper explores the potential of incorporating quantum-like phenomena into learning algorithms to mitigate noise and facilitate adaptation. We propose a novel framework that leverages superposition and entanglement to create models less susceptible to fluctuations in input data, ultimately improving predictive performance. The core mechanism focuses on introducing these quantum-inspired elements to promote stability and resilience in model training and inference. The research aims to address limitations of current machine learning approaches by introducing a fundamentally different method for handling uncertainty and exploring robust solutions. We present preliminary results demonstrating the effectiveness of this framework in several benchmark datasets, highlighting its potential for advancing the field.

Sep 6, 20260 citations

Quantum-Enhanced Gradient Descent for Deep Learning

Jincheng Zhang

Quantum-enhanced gradient descent offers a promising approach to accelerate and improve the training of deep learning models. This research investigates the potential of integrating quantum annealing techniques into the gradient descent optimization process, aiming to overcome limitations of classical methods in navigating complex loss landscapes. The core claim is that quantum annealing can significantly enhance convergence speed and robustness, particularly in deep learning scenarios. The paper presents a theoretical analysis and preliminary implementation demonstrating the effectiveness of this approach, highlighting the potential benefits and challenges of this novel methodology.

Sep 7, 20260 citations

Quantum-Inspired Stochastic Gradient Descent with Entangled Parameter Updates

Jincheng Zhang

This paper proposes a novel approach to stochastic gradient descent (SGD) inspired by quantum entanglement. The core idea is that parameter updates in a neural network can be significantly accelerated by leveraging correlations between parameters, mimicking the behavior of entangled quantum systems. We introduce a shared tensor network representing a quantum state amongst the network's parameters. Parameter updates are then performed by manipulating this shared state, effectively creating an entangled state that reduces the variance of the gradient estimate. This approach, termed Quantum-Inspired SGD with Entangled Parameter Updates (QISGD-EPU), offers a potential pathway to dramatically improve the training speed of deep neural networks. Theoretical analysis suggests that the correlation inherent in the shared state can lead to faster convergence and reduced sensitivity to the learning rate. We outline the mathematical framework, the update rules, and discuss the potential benefits and challenges of this new paradigm.

Sep 6, 20260 citations