PulseAugur
EN
LIVE 17:04:38

New federated learning methods tackle efficiency and generalization

By PulseAugur Editorial · [4 sources] ·

Researchers have developed new methods for federated learning, a technique that allows decentralized agents to collaboratively train models without sharing raw data. One approach, FedQHD, uses a specific structure for reinforcement learning agents to enable closed-form aggregation of parameters, improving efficiency and performance on control benchmarks. Another paper introduces a framework called FedDTL for federated vision-language models, which decouples image and text encoders to reduce inconsistencies and uses a two-stage fine-tuning process involving reinforcement learning for better generalization. Additionally, a method named C-MOPPO addresses joint optimization of training and inference in federated edge learning by formulating it as a constrained multi-objective Markov decision process, balancing accuracy, latency, and energy consumption. AI

IMPACT These advancements in federated learning offer improved efficiency, generalization, and resource management for decentralized AI systems.

RANK_REASON Cluster contains multiple academic papers detailing novel research methodologies in federated learning.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 4 sources. How we write summaries →

New federated learning methods tackle efficiency and generalization

COVERAGE [4]

  1. arXiv cs.LG TIER_1 English(EN) · Yuchen Hou, Yongshan Chen, Zhuowen Zou, Calvin Yeung, Mohsen Imani, Tian Lan, Mahdi Imani ·

    FedQHD: Closed-Form Function-Space Federated Reinforcement Learning

    arXiv:2605.29002v1 Announce Type: new Abstract: Federated reinforcement learning enables decentralized agents to collaboratively improve policies or value estimates without exchanging raw trajectories. However, FedAvg-style parameter averaging is not function-space consistent: wh…

  2. arXiv cs.LG TIER_1 English(EN) · Zhen Li, Jun Cai, Chao Yang, Haoran Gao ·

    Joint Optimization of Training and Inference in Federated Edge Learning via Constrained Multi-Objective Deep Reinforcement Learning

    arXiv:2605.25916v1 Announce Type: new Abstract: Federated edge learning (FEEL) has recently emerged as a promising paradigm for achieving edge intelligence (EI) via enabling collaborative model training across edge devices while protecting data privacy. In this paper, we put fort…

  3. arXiv cs.LG TIER_1 English(EN) · Haoran Gao ·

    Joint Optimization of Training and Inference in Federated Edge Learning via Constrained Multi-Objective Deep Reinforcement Learning

    Federated edge learning (FEEL) has recently emerged as a promising paradigm for achieving edge intelligence (EI) via enabling collaborative model training across edge devices while protecting data privacy. In this paper, we put forth an online optimization framework that jointly …

  4. arXiv cs.CV TIER_1 English(EN) · Yuting Ma, Lechao Cheng, Xiaohua Xu ·

    Decoupled Training with Local Reinforcement Fine-Tuning in Federated Learning

    arXiv:2605.27900v1 Announce Type: new Abstract: Federated Learning (FL) with pre-trained Vision-Language Models (VLMs) has emerged as a promising paradigm for various downstream tasks. By leveraging its strong representations, recent studies improve task adaptation under insuffic…

RELATED ENTITIES

RELATED TOPICS