A Drift Stable Quantum Federated Learning for Intelligent Services

Short summary

The authors propose DUQFL-Prox, a drift-stable quantum federated learning framework that uses deep-unfolded local optimization with adaptive SPSA updates and a proximal term to keep local models close to the global model. A lightweight controller learns step-specific optimization parameters to improve post-aggregation performance. Experiments on financial fraud and genomic classification tasks show improvements in stability, generalization, and client fairness compared to standard quantum federated learning baselines.

• •DUQFL-Prox combines deep-unfolded SPSA with proximal regularization to reduce client drift in quantum federated learning
• •Lightweight controller learns per-step optimization parameters for better aggregation
• •Experiments on fraud detection and genomic classification show improved stability and fairness over QFL baselines

Generated with AI, which can make mistakes.