Results 51 to 60 of about 7,049,980 (289)

Balancing Privacy and Performance: A Differential Privacy Approach in Federated Learning

Computers
Federated learning (FL), a decentralized approach to machine learning, facilitates model training across multiple devices, ensuring data privacy. However, achieving a delicate privacy preservation–model convergence balance remains a major problem ...
Huda Kadhim Tayyeh, Ahmed Sabah Ahmed AL-Jumaili   +1 more
doaj   +1 more source

Ensemble Federated Adversarial Training with Non-IID data

CoRR, 2021
Despite federated learning endows distributed clients with a cooperative training mode under the premise of protecting data privacy and security, the clients are still vulnerable when encountering adversarial samples due to the lack of robustness.
Shuang Luo, Didi Zhu, Zexi Li 0001, Chao Wu 0001   +3 more
openaire   +2 more sources

Factor-of-iid balanced orientation of non-amenable graphs

, 2021
We show that if a non-amenable, quasi-transitive, unimodular graph $G$ has all degrees even then it has a factor-of-iid balanced orientation, meaning each vertex has equal in- and outdegree.
Hrušková, Aranka, Tóth, László Márton, Bencs, Ferenc   +2 more
core   +4 more sources

Global Layers: Non-IID Tabular Federated Learning

CoRR, 2023
Data heterogeneity between clients remains a key challenge in Federated Learning (FL), particularly in the case of tabular data. This work presents Global Layers (GL), a novel partial model personalization method robust in the presence of joint distribution $P(X,Y)$ shift and mixed input/output spaces $X \times Y$ across clients.
openaire   +2 more sources

New‐Era Polymer Thermoelectrics: Material Innovations, Doping Frontiers, Decoupling Strategies, and Unconventional Applications

Advanced Materials, EarlyView.
The field of polymer thermoelectrics is entering a new era, featuring breakthroughs in addressing the conventional performance disparity between p‐type and n‐type polymers, pioneering doping frontiers, and sophisticated decoupling strategies. This review explores innovations in molecular design and superior stabilities, bridging the gap from ...
Suhao Wang
wiley   +1 more source

Beyond IID: Learning to combine Non-IID metrics for vision tasks

, 2017
Copyright © 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. Metric learning has been widely employed, especially in various computer vision tasks, with the fundamental assumption that all samples (e.g.
Gao, Y, Shi, Yinghuan, Gao, Yang, Li, W, Cao, L, Shen, Dinggang, Li, Wenbin, Shi, Y, Shen, D, Cao, Longbing   +9 more
core   +1 more source

Continual Learning for Multimodal Data Fusion of a Soft Gripper

Advanced Robotics Research, EarlyView.
Models trained on a single data modality often struggle to generalize when exposed to a different modality. This work introduces a continual learning algorithm capable of incrementally learning different data modalities by leveraging both class‐incremental and domain‐incremental learning scenarios in an artificial environment where labeled data is ...
Nilay Kushawaha, Egidio Falotico
wiley   +1 more source

The KIF6‐RBP Complex Orchestrates mRNA Transport Required for Sperm Flagellar Assembly

Advanced Science, EarlyView.
Two homozygous deleterious KIF6 variants are identified in unrelated men with impaired sperm motility. Mouse models and multi‐omics analyses reveal that KIF6 cooperates with the RNA‐binding proteins FMRP and FXR1 to deliver mRNAs essential for sperm flagellar assembly, linking disrupted mRNA transport to reduced abundance of key structural and ...
Chunbo Xie, Sibing Yi, Xinle Lin, Shen Zhang, Weili Wang, Shimin Yuan, Lanlan Meng, Yong Li, Chen Tan, Chunjia Wei, Yanyan Yu, Yaoqiong Liang, Huan Zhang, Liang Hu, Guangxiu Lu, Wenbin He, Qianjun Zhang, Juan Du, Ge Lin, Chaofeng Tu, Yue‐Qiu Tan   +20 more
wiley   +1 more source

RRAM Variability Harvesting for CIM‐Integrated TRNG

Advanced Electronic Materials, EarlyView.
This work demonstrates a compute‐in‐memory‐compatible true random number generator that harvests intrinsic cycle‐to‐cycle variability from a 1T1R RRAM array. Parallel entropy extraction enables high‐throughput bit generation without dedicated circuits. This approach achieves NIST‐compliant randomness and low per‐bit energy, offering a scalable hardware
Ankit Bende, Gokulnath Rajendran, Regina Dittmann, Anupam Chattopadhyay, Vikas Rana   +4 more
wiley   +1 more source

Explainable Federated Learning for Brain Tumor Classification Using Multi-Source MRI Data

Iraqi Journal for Computers and Informatics
Early diagnosis and clinical decision-making depend on accurate brain tumor classification using magnetic resonance imaging (MRI). However, traditional deep learning methods usually rely on centralized medical data, which raises privacy concerns and ...
Suhad Muhy Helal, Belal Al-Khateeb
doaj   +1 more source