Federated Learning with Efficient Local Adaptation for Realized Volatility Prediction
Abstract: Financial markets present unique challenges for Federated Learning (FL) due to fragmented datasets, dynamic participation, and the critical need for precise and reliable predictions. Isolated local datasets often fail to capture the full spectrum of market dynamics, blocking accurate realized volatility predictions. Unlike traditional FL methods that focus on improving convergence during the training process, we propose Federated Learning with Adaptive Robustness and Efficiency for Local Adaptation (FLARE-LA), a novel framework designed to optimize predictive performance after the global training phase. FLARE-LA leverages Taylor-based local linearization and probabilistic optimization to efficiently adapt global models to local data distributions, enabling fast responsiveness to new market conditions. This adaptability ensures trained local models align with real-world scenarios, making FLARE-LA particularly suited to dynamic financial applications. Extensive experimental evaluations demonstrate FLARE-LA's superior performance, showcasing its ability to significantly enhance post-FL outcomes compared to state-of-the-art FL algorithms. The results underscore FLARE-LA's unique capability to drive advancements in financial forecasting and other high-stakes, rapidly evolving domains.
Submission Length: Long submission (more than 12 pages of main content)
Changes Since Last Submission: 1. Implemented and tested the requested baselines using extra local updates after the FL training procedure. Results have been added to Section 5.3, demonstrating the advantages of FLARE-LA over the new enhanced baselines in terms of robustness and generalization.
2. Based on feedback, Appendix A has been removed. Instead, we focused on enhancing Section 4.2 to provide a detailed, structured, and rigorous analysis of the robust local adaptation mechanism within FLARE-LA.
3. Numerical results have been removed from the abstract to improve clarity and relevance as suggested.
4. Section 4.1 now includes a pseudocode representation of the FLARE-LA algorithm, along with detailed descriptions of the required hyperparameters.
5. The paper has been revised for conciseness and clarity. Section 4.2 were rewritten to focus on essential analyses. Non-critical sections, e.g., parts of Section 4.2, were removed as suggested.
6. Section 4.1 now addresses different settings of conditioning, including cases where the matrix is not invertible, to ensure robustness and completeness.
7. All inline citations were updated to align with the format (author, year) unless the author's name is part of the sentence.
Assigned Action Editor: ~Zhiyu_Zhang1
Submission Number: 3393
Loading