Researchers have introduced Federated Integrative Two-Step Synthesis (FITS), a novel framework designed for generating mixed-type tabular synthetic data within federated learning environments. FITS addresses the challenges of maintaining distributional fidelity and effectively managing diverse data types by employing a two-step federated algorithm that estimates conditional cumulative distribution functions using local linear regression. The framework incorporates a continualization method to unify discrete and continuous variables, enabling more robust synthetic data generation. Experiments indicate that FITS-generated data enhances downstream task performance for all participating clients. AI
IMPACT Enhances downstream task performance in federated learning settings through improved synthetic data generation.
RANK_REASON The cluster contains a research paper detailing a new method for synthetic data generation in federated learning. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Fits Interpretable Spiking Neurons
- Gotit.pub
- Hugging Face
- IArxiv
- Influence Flower
- ScienceCast
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
