Back to feed
arXiv cs.LG
arXiv cs.LG
7/7/2026
QuantFlow: A Federated Mamba-Based Post-Transformer Foundation Model for Time-Series Forecasting

QuantFlow: A Federated Mamba-Based Post-Transformer Foundation Model for Time-Series Forecasting

Short summary

QuantFlow combines Mamba state-space networks with federated learning to enable privacy-preserving time-series forecasting without centralizing data. The model achieves competitive accuracy across crypto, weather, electricity, and traffic domains while supporting decentralized deployment with minimal communication. Trade-offs include reduced performance on long-horizon forecasts and irregular epidemiological signals.

  • •New architecture combining Mamba networks with federated learning for distributed forecasting
  • •Strong empirical results on crypto, weather, electricity, and traffic datasets with privacy preservation
  • •Handles decentralized deployment with non-IID data over 20 clients in 3 communication rounds

Generated with AI, which can make mistakes.

#research-breakthrough#ai-tools
Read full article at arXiv cs.LG

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

STKAN: Kolmogorov-Arnold Networks for Spatio-Temporal Forecasting
arXiv cs.LG
HantaWatch: Federated Learning for Hantavirus Genomic Surveillance
arXiv cs.LG
NEST: Tackling Dataset-Level Distribution Shifts via Regime-Oriented Mixture-of-Experts
arXiv cs.LG
When to Trust, How to Distill: Multi-Foundation Model Guidance for Lightweight, Robust Scientific Time Series Forecasting
arXiv cs.LG