Federated Quantum Machine Learning.

Area of Science:

• Quantum Machine Learning
• Distributed Computing
• Artificial Intelligence

Background:

• Federated learning (FL) offers a privacy-preserving approach to distributed machine learning by training models locally on decentralized data.
• Current quantum machine learning (QML) research has not explored federated settings, leaving potential benefits for privacy and scalability unaddressed.

Purpose of the Study:

• To introduce and evaluate a federated learning framework for hybrid quantum-classical machine learning models.
• To investigate the potential of federated learning in accelerating QML training and enhancing data privacy.

Main Methods:

• Developed a federated learning scheme for distributed training of hybrid quantum-classical models.
• Utilized quantum neural networks (QNNs) integrated with pre-trained classical convolutional models.
• Evaluated the performance in terms of model accuracy and training time compared to centralized approaches.

Main Results:

• The proposed federated learning scheme achieved trained model accuracies comparable to centralized methods.
• Demonstrated a significant reduction in distributed training time using the federated approach.
• Validated the framework's applicability to hybrid QML models, with potential for generalization to pure QML.

Conclusions:

• Federated learning is a viable and effective approach for training hybrid quantum-classical machine learning models.
• This work opens promising research avenues for scaling QML and improving data privacy in distributed settings.