Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantum-Evolutionary Neural Networks for Multi-Agent Federated Learning

Aarav Lala, Kalyan Cherukuri

TL;DR

Addressing privacy-preserving real-time decision-making in decentralized multi-agent systems, the paper blends quantum-inspired neural networks with evolutionary optimization and federated learning to enable adaptive, privacy-safe collaboration. The QE-NN employs phase-shift sine activations to mimic quantum interference and an evolutionary local search to refine client models, with aggregation protected by Gaussian noise to guarantee privacy. The authors provide convergence guarantees under standard smoothness and bounded-variance assumptions and $(\epsilon,\delta)$-DP guarantees for updates, while experiments on synthetic data and benchmarks indicate global accuracy near $0.97$ and resilience to data heterogeneity. This framework advances privacy-aware, quantum-inspired distributed learning with potential applications in autonomous systems, smart cities, and healthcare.

Abstract

As artificial intelligence continues to drive innovation in complex, decentralized environments, the need for scalable, adaptive, and privacy-preserving decision-making systems has become critical. This paper introduces a novel framework combining quantum-inspired neural networks with evolutionary algorithms to optimize real-time decision-making in multi-agent systems (MAS). The proposed Quantum-Evolutionary Neural Network (QE-NN) leverages quantum computing principles -- such as quantum superposition and entanglement -- to enhance learning speed and decision accuracy, while integrating evolutionary optimization to continually refine agent behaviors in dynamic, uncertain environments. By utilizing federated learning, QE-NN ensures privacy preservation, enabling decentralized agents to collaborate without sharing sensitive data. The framework is designed to allow agents to adapt in real-time to their environments, optimizing decision-making processes for applications in areas such as autonomous systems, smart cities, and healthcare. This research represents a breakthrough in merging quantum computing, evolutionary optimization, and privacy-preserving techniques to solve complex problems in multi-agent decision-making systems, pushing the boundaries of AI in real-world, privacy-sensitive applications.

Quantum-Evolutionary Neural Networks for Multi-Agent Federated Learning

TL;DR

Addressing privacy-preserving real-time decision-making in decentralized multi-agent systems, the paper blends quantum-inspired neural networks with evolutionary optimization and federated learning to enable adaptive, privacy-safe collaboration. The QE-NN employs phase-shift sine activations to mimic quantum interference and an evolutionary local search to refine client models, with aggregation protected by Gaussian noise to guarantee privacy. The authors provide convergence guarantees under standard smoothness and bounded-variance assumptions and -DP guarantees for updates, while experiments on synthetic data and benchmarks indicate global accuracy near and resilience to data heterogeneity. This framework advances privacy-aware, quantum-inspired distributed learning with potential applications in autonomous systems, smart cities, and healthcare.

Abstract

As artificial intelligence continues to drive innovation in complex, decentralized environments, the need for scalable, adaptive, and privacy-preserving decision-making systems has become critical. This paper introduces a novel framework combining quantum-inspired neural networks with evolutionary algorithms to optimize real-time decision-making in multi-agent systems (MAS). The proposed Quantum-Evolutionary Neural Network (QE-NN) leverages quantum computing principles -- such as quantum superposition and entanglement -- to enhance learning speed and decision accuracy, while integrating evolutionary optimization to continually refine agent behaviors in dynamic, uncertain environments. By utilizing federated learning, QE-NN ensures privacy preservation, enabling decentralized agents to collaborate without sharing sensitive data. The framework is designed to allow agents to adapt in real-time to their environments, optimizing decision-making processes for applications in areas such as autonomous systems, smart cities, and healthcare. This research represents a breakthrough in merging quantum computing, evolutionary optimization, and privacy-preserving techniques to solve complex problems in multi-agent decision-making systems, pushing the boundaries of AI in real-world, privacy-sensitive applications.

Paper Structure

This paper contains 21 sections, 14 equations, 4 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Federated Learning under Non-IID Data Distributions
  4. Evolutionary Algorithms in Distributed Optimization
  5. Quantum-Inspired Neural Architectures and Privacy Mechanisms
  6. Problem Formulation
  7. Federated Learning Objective
  8. Quantum-Evolutionary Neural Network (QE-NN)
  9. Local Evolutionary Optimization
  10. Privacy-Preserving Aggregation
  11. Final Optimization Objective
  12. Theoretical Validation
  13. Proof of Convergence
  14. Superposition and Entanglement
  15. Evolutionary Benefit & Privacy Guarantee
  16. ...and 6 more sections

Figures (4)

  • Figure 1: From superposition to entanglement: A single qubit $\ket{\psi}$ can be extended into an entangled two-qubit state $\alpha \ket{00} + \beta \ket{11}$.
  • Figure 2: Quantum-Evolutionary Federated Learning Pipeline
  • Figure 3: Accuracy of Global Model on Synthetic Dataset over Training Rounds
  • Figure 4: Performance comparison across MNIST, CIFAR10, and CIFAR100 for accuracy, F1 score, and loss.