AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

EvoFed: Leveraging Evolutionary Strategies for Communication-Efficient Federated Learning

Mohammad Mahdi Rahimi, Hasnain Irshad Bhatti, Younghyun Park, Humaira Kousar, Do-Yeon Kim, Jaekyun Moon

NeurIPS 2023 · Federated learning and privacy › Communication-efficient FL · approach: Efficient local training in FL · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Replace model-parameter exchange with PBGE fitness-vector exchangeFedAvg-style federated learning · MNISTcommunication cost at max accuracy: 73.7 -> 9.2 MBImproved
Replace model-parameter exchange with PBGE fitness-vector exchangeFedAvg-style federated learning · FMNISTcommunication cost at max accuracy: 40.99 -> 7.78 MBImproved
Replace model-parameter exchange with PBGE fitness-vector exchangeFedAvg-style federated learning · CIFAR-10accuracy: 50.22% -> 54.12%Improved
Replace model-parameter exchange with PBGE fitness-vector exchangeFedAvg-style federated learning · CIFAR-10communication cost at max accuracy: 2.134 -> 0.023 GBImproved
Use a larger initial population (ablation)EvoFed · FMNISTaccuracyImproved
Increase population size beyond 128 (ablation)EvoFed · FMNISTaccuracyNo clear effect
Increase the number of clients above 5 (ablation)EvoFed · image classification experimentsaccuracyHurt
Sparsify the population to fewer remaining members (ablation)EvoFed · image classificationperformanceHurt
Start with a larger population, then sparsify to a smaller set (ablation)EvoFed · image classificationperformanceImproved
Quantize fitness values, including to 2 bits (ablation)EvoFed · image classificationperformanceHurt
Transmit fitness ranks instead of fitness values (ablation)EvoFed · image classificationperformanceImproved
Use quantization rather than ranking at the same bit budget (ablation)EvoFed · image classificationperformanceImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy