Saved in:
Bibliographic Details
Main Authors: Sharma, Karan, Tripathi, Aditya, Mishra, Rahul, Maiti, Tapas Kumar
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.24418
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910250978770944
author Sharma, Karan
Tripathi, Aditya
Mishra, Rahul
Maiti, Tapas Kumar
author_facet Sharma, Karan
Tripathi, Aditya
Mishra, Rahul
Maiti, Tapas Kumar
contents Federated learning is used in medical imaging where privacy prohibits centralizing data. Standard federated algorithms assume homogeneous hardware, identical architectures, and centralized aggregation, which fails when hospitals have unequal compute resources. We propose capacity-aware coordination: measure each hospital's throughput, assign capacity-appropriate architectures (MobileNetV3-Small, EfficientNet-B0, ResNet-50), and combine predictions via weighted ensemble. Weak and strong hospitals can participate without forcing uniform architectures. We separate on-chain policy from off-chain learning. A Solidity contract stores hospital registration, benchmark hashes, metrics, and weights. Hospitals train locally and submit only hashes and scalars (not parameters). Weighted ensemble inference is computed off-chain. Experiments on PneumoniaMNIST and DermaMNIST (5 seeds, 3 non-IID levels) show our method achieves lower or equal calibration error versus equal-weight ensemble and competitive accuracy versus FedAvg, FedProx, and FedMD. Communication overhead is 224 bytes per round, a reduction of over 912,000x compared to FedAvg.
format Preprint
id arxiv_https___arxiv_org_abs_2605_24418
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle ChainLearn: A Blockchain-Based Capacity-Aware Framework for Federated Ensemble Learning
Sharma, Karan
Tripathi, Aditya
Mishra, Rahul
Maiti, Tapas Kumar
Machine Learning
Federated learning is used in medical imaging where privacy prohibits centralizing data. Standard federated algorithms assume homogeneous hardware, identical architectures, and centralized aggregation, which fails when hospitals have unequal compute resources. We propose capacity-aware coordination: measure each hospital's throughput, assign capacity-appropriate architectures (MobileNetV3-Small, EfficientNet-B0, ResNet-50), and combine predictions via weighted ensemble. Weak and strong hospitals can participate without forcing uniform architectures. We separate on-chain policy from off-chain learning. A Solidity contract stores hospital registration, benchmark hashes, metrics, and weights. Hospitals train locally and submit only hashes and scalars (not parameters). Weighted ensemble inference is computed off-chain. Experiments on PneumoniaMNIST and DermaMNIST (5 seeds, 3 non-IID levels) show our method achieves lower or equal calibration error versus equal-weight ensemble and competitive accuracy versus FedAvg, FedProx, and FedMD. Communication overhead is 224 bytes per round, a reduction of over 912,000x compared to FedAvg.
title ChainLearn: A Blockchain-Based Capacity-Aware Framework for Federated Ensemble Learning
topic Machine Learning
url https://arxiv.org/abs/2605.24418