Federated Split Learning with Only Positive Labels for resource-constrained IoT environment

TL;DR

This paper introduces SFPL, a novel federated split learning method tailored for resource-limited IoT devices with only positive labels, significantly improving classification performance in distributed deep learning scenarios.

Contribution

The paper proposes SFPL, a new approach that enhances federated split learning for positive-only labeled data in IoT, addressing convergence issues and boosting accuracy.

Findings

SFPL outperforms SFL by large factors on CIFAR datasets.

SFPL effectively handles positive-only labeled data in resource-constrained IoT devices.

SFPL demonstrates significant accuracy improvements in distributed deep learning.

Abstract

Distributed collaborative machine learning (DCML) is a promising method in the Internet of Things (IoT) domain for training deep learning models, as data is distributed across multiple devices. A key advantage of this approach is that it improves data privacy by removing the necessity for the centralized aggregation of raw data but also empowers IoT devices with low computational power. Among various techniques in a DCML framework, federated split learning, known as splitfed learning (SFL), is the most suitable for efficient training and testing when devices have limited computational capabilities. Nevertheless, when resource-constrained IoT devices have only positive labeled data, multiclass classification deep learning models in SFL fail to converge or provide suboptimal results. To overcome these challenges, we propose splitfed learning with positive labels (SFPL). SFPL applies a random shuffling function to the smashed data received from clients before supplying it to the server for model training. Additionally, SFPL incorporates the local batch normalization for the client-side model portion during the inference phase. Our results demonstrate that SFPL outperforms SFL: (i) by factors of 51.54 and 32.57 for ResNet-56 and ResNet-32, respectively, with the CIFAR-100 dataset, and (ii) by factors of 9.23 and 8.52 for ResNet-32 and ResNet-8, respectively, with CIFAR-10 dataset. Overall, this investigation underscores the efficacy of the proposed SFPL framework in DCML.