GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → litian96 → FedDANE

litian96 / FedDANE

Licence: MIT license
FedDANE: A Federated Newton-Type Method (Asilomar Conference on Signals, Systems, and Computers ‘19)

Programming Languages

python
139335 projects - #7 most used programming language
shell
77523 projects

Labels

distributed-optimizationfederated-learning

Projects that are alternatives of or similar to FedDANE

MOON
Model-Contrastive Federated Learning (CVPR 2021)
Stars: ✭ 93 (+272%)
Mutual labels:  federated-learning
Awesome-Federated-Machine-Learning
Everything about federated learning, including research papers, books, codes, tutorials, videos and beyond
Stars: ✭ 190 (+660%)
Mutual labels:  federated-learning
Awesome Mlops
A curated list of references for MLOps
Stars: ✭ 7,119 (+28376%)
Mutual labels:  federated-learning
backdoors101
Backdoors Framework for Deep Learning and Federated Learning. A light-weight tool to conduct your research on backdoors.
Stars: ✭ 181 (+624%)
Mutual labels:  federated-learning
easyFL
An experimental platform to quickly realize and compare with popular centralized federated learning algorithms. A realization of federated learning algorithm on fairness (FedFV, Federated Learning with Fair Averaging, https://fanxlxmu.github.io/publication/ijcai2021/) was accepted by IJCAI-21 (https://www.ijcai.org/proceedings/2021/223).
Stars: ✭ 104 (+316%)
Mutual labels:  federated-learning
federated-xgboost
Federated gradient boosted decision tree learning
Stars: ✭ 39 (+56%)
Mutual labels:  federated-learning
fedpa
Federated posterior averaging implemented in JAX
Stars: ✭ 38 (+52%)
Mutual labels:  federated-learning
FedDA
Source code for 'Dual Attention Based FL for Wireless Traffic Prediction'
Stars: ✭ 41 (+64%)
Mutual labels:  federated-learning
FATE-Serving
A scalable, high-performance serving system for federated learning models
Stars: ✭ 107 (+328%)
Mutual labels:  federated-learning
Fate
An Industrial Grade Federated Learning Framework
Stars: ✭ 3,775 (+15000%)
Mutual labels:  federated-learning
FedLab-benchmarks
Standard federated learning implementations in FedLab and FL benchmarks.
Stars: ✭ 49 (+96%)
Mutual labels:  federated-learning
Federated-Learning-Mini-Framework
Federated Learning mini-framework with Keras
Stars: ✭ 38 (+52%)
Mutual labels:  federated-learning
Challenge
The repo for the FeTS Challenge
Stars: ✭ 21 (-16%)
Mutual labels:  federated-learning
KD3A
Here is the official implementation of the model KD3A in paper "KD3A: Unsupervised Multi-Source Decentralized Domain Adaptation via Knowledge Distillation".
Stars: ✭ 63 (+152%)
Mutual labels:  federated-learning
Pysyft
A library for answering questions using data you cannot see
Stars: ✭ 7,811 (+31144%)
Mutual labels:  federated-learning
FedReID
Implementation of Federated Learning to Person Re-identification (Code for ACMMM 2020 paper)
Stars: ✭ 68 (+172%)
Mutual labels:  federated-learning
Front-End
Federated Learning based Deep Learning. Docs: https://fets-ai.github.io/Front-End/
Stars: ✭ 35 (+40%)
Mutual labels:  federated-learning
DeML-Golem
Proof Of Concept of DEcentralised Machine Learning on top of the Golem (https://golem.network/) architecture
Stars: ✭ 35 (+40%)
Mutual labels:  federated-learning
shadho
Scalable, structured, dynamically-scheduled hyperparameter optimization.
Stars: ✭ 17 (-32%)
Mutual labels:  distributed-optimization
baai-federated-learning-crane-baseline
电力人工智能数据竞赛——液压吊车目标检测赛道
Stars: ✭ 17 (-32%)
Mutual labels:  federated-learning
View All Similar Projects ➔

FedDANE: A Federated Newton-Type Method

This repository contains the code and experiments for the paper:

FedDANE: A Federated Newton-Type Method

Asilomar Conference on Signals, Systems, and Computers 2019 (Invited Paper)

FedDANE is an optimization method that we adapt from DANE, a method for classical distributed optimization, to handle the practical constraints of federated learning. We provide convergence guarantees for this method when learning over both convex and non-convex functions. Despite encouraging theoretical results, we find that the method has underwhelming performance empirically. We identify low device participation and statistical device heterogeneity as two underlying causes of this underwhelming performance. In the paper, we also suggest several directions of future work.

This repository contains a set of empirical evaluation on both synthetic and real-world datasets. FedDANE consistently underperforms baselines of FedAvg and FedProx in realistic federated settings.

Usage

The usage is almost the same as that of the FedProx code, except that we are using a different optimizer specified in flearn/optimizer/pggd.py. For example, command lines to reproduce the results on the synthetic IID data are:

mkdir log_synthetic
bash run_trainer.sh synthetic_iid fedavg 0 | tee log_synthetic/iid_fedavg_c10_e20
bash run_trainer.sh synthetic_iid fedprox 1 | tee log_synthetic/iid_fedprox_c10_e20
bash run_trainer.sh synthetic_iid feddane 0 | tee log_synthetic/iid_feddane_c10_e20
python plot.py

References

See our FedDANE paper for more details as well as all references.

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].