Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data

Federated learning is increasingly being considered for sensor-driven human activity recognition, offering advantages in terms of privacy and scalability compared to centralized methods. However, challenges such as feature selection and client imbalanced data persist. In this study, FLP-DS2MOTE-USA...

Full description

Bibliographic Details
Main Authors: Taha, Zahraa Khduair, Siaw Paw, Johnny Koh, Tak, Yaw Chong, Kiong, Tiong Sieh, Kadirgama, Kumaran, Benedict, Foo, Ding, Tan Jian, Kharudin, Ali, Abed, Azher M.
Format: Article
Language:English
Published: IEEE 2024
Subjects:
Online Access:http://umpir.ump.edu.my/id/eprint/44790/
http://umpir.ump.edu.my/id/eprint/44790/1/Advances%20in%20federated%20learning-Combining%20local%20preprocessing%20with%20adaptive.pdf
_version_ 1848827183450030080
author Taha, Zahraa Khduair
Siaw Paw, Johnny Koh
Tak, Yaw Chong
Kiong, Tiong Sieh
Kadirgama, Kumaran
Benedict, Foo
Ding, Tan Jian
Kharudin, Ali
Abed, Azher M.
author_facet Taha, Zahraa Khduair
Siaw Paw, Johnny Koh
Tak, Yaw Chong
Kiong, Tiong Sieh
Kadirgama, Kumaran
Benedict, Foo
Ding, Tan Jian
Kharudin, Ali
Abed, Azher M.
author_sort Taha, Zahraa Khduair
building UMP Institutional Repository
collection Online Access
description Federated learning is increasingly being considered for sensor-driven human activity recognition, offering advantages in terms of privacy and scalability compared to centralized methods. However, challenges such as feature selection and client imbalanced data persist. In this study, FLP-DS2MOTE-USA is suggested, a system that integrates federated local preprocessing, adaptive thresholding based on uncertainty symmetry, and a density- sensitive synthetic minority over-sampling approach. Each client preprocesses data locally and employs DS2MOTE for class balancing. On the server side, adaptive thresholding based on uncertainty symmetry is utilized to identify the optimal client for training the global mode. Evaluation on two distinct datasets—Human Activity Recognition with Smartphones and Human Activity Recognition (OpenPose) —reveals that our model outperforms FedAvg, FedSgd, FedSmote, and FedNova, achieving accuracies of 90.57% and 96.58%, respectively. In addition, FLP-DS2MOTE-USA minimizes update size and network overhead on the Human Activity Recognition with Smartphones, while achieving improvements on the OpenPose dataset. Overall, the proposed method not only addresses issues of imbalanced data but also reduces computational complexity via streamlined local preprocessing, and server-side mechanisms ensure client privacy. It outperforms traditional federated learning techniques in both accuracy and efficiency.
first_indexed 2025-11-15T03:56:40Z
format Article
id ump-44790
institution Universiti Malaysia Pahang
institution_category Local University
language English
last_indexed 2025-11-15T03:56:40Z
publishDate 2024
publisher IEEE
recordtype eprints
repository_type Digital Repository
spelling ump-447902025-06-12T07:21:00Z http://umpir.ump.edu.my/id/eprint/44790/ Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data Taha, Zahraa Khduair Siaw Paw, Johnny Koh Tak, Yaw Chong Kiong, Tiong Sieh Kadirgama, Kumaran Benedict, Foo Ding, Tan Jian Kharudin, Ali Abed, Azher M. QA75 Electronic computers. Computer science TJ Mechanical engineering and machinery Federated learning is increasingly being considered for sensor-driven human activity recognition, offering advantages in terms of privacy and scalability compared to centralized methods. However, challenges such as feature selection and client imbalanced data persist. In this study, FLP-DS2MOTE-USA is suggested, a system that integrates federated local preprocessing, adaptive thresholding based on uncertainty symmetry, and a density- sensitive synthetic minority over-sampling approach. Each client preprocesses data locally and employs DS2MOTE for class balancing. On the server side, adaptive thresholding based on uncertainty symmetry is utilized to identify the optimal client for training the global mode. Evaluation on two distinct datasets—Human Activity Recognition with Smartphones and Human Activity Recognition (OpenPose) —reveals that our model outperforms FedAvg, FedSgd, FedSmote, and FedNova, achieving accuracies of 90.57% and 96.58%, respectively. In addition, FLP-DS2MOTE-USA minimizes update size and network overhead on the Human Activity Recognition with Smartphones, while achieving improvements on the OpenPose dataset. Overall, the proposed method not only addresses issues of imbalanced data but also reduces computational complexity via streamlined local preprocessing, and server-side mechanisms ensure client privacy. It outperforms traditional federated learning techniques in both accuracy and efficiency. IEEE 2024 Article PeerReviewed pdf en cc_by_nc_nd_4 http://umpir.ump.edu.my/id/eprint/44790/1/Advances%20in%20federated%20learning-Combining%20local%20preprocessing%20with%20adaptive.pdf Taha, Zahraa Khduair and Siaw Paw, Johnny Koh and Tak, Yaw Chong and Kiong, Tiong Sieh and Kadirgama, Kumaran and Benedict, Foo and Ding, Tan Jian and Kharudin, Ali and Abed, Azher M. (2024) Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data. IEEE Access, 12. pp. 186277-186295. ISSN 2169-3536. (Published) http://doi.org/10.1109/ACCESS.2024.3435910 http://doi.org/10.1109/ACCESS.2024.3435910
spellingShingle QA75 Electronic computers. Computer science
TJ Mechanical engineering and machinery
Taha, Zahraa Khduair
Siaw Paw, Johnny Koh
Tak, Yaw Chong
Kiong, Tiong Sieh
Kadirgama, Kumaran
Benedict, Foo
Ding, Tan Jian
Kharudin, Ali
Abed, Azher M.
Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title_full Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title_fullStr Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title_full_unstemmed Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title_short Advances in federated learning: Combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
title_sort advances in federated learning: combining local preprocessing with adaptive uncertainty symmetry to reduce irrelevant features and address imbalanced data
topic QA75 Electronic computers. Computer science
TJ Mechanical engineering and machinery
url http://umpir.ump.edu.my/id/eprint/44790/
http://umpir.ump.edu.my/id/eprint/44790/
http://umpir.ump.edu.my/id/eprint/44790/
http://umpir.ump.edu.my/id/eprint/44790/1/Advances%20in%20federated%20learning-Combining%20local%20preprocessing%20with%20adaptive.pdf