Jadwal Sholat

Memuat jadwal sholatโ€ฆ

Ilmu Komputer & AI editorial

Open AccessOA2023

Fused Weighted Federated Deep Extreme Machine Learning Based on Intelligent Lung Cancer Disease Prediction Model for Healthcare 5.0

This paper proposes a fused weighted federated deep extreme learning model for lung cancer prediction in Healthcare 5.0, addressing data privacy and achieving 97.2% accuracy, outperforming state-of-the-art methods.
Sagheer Abbas; G. Issa; Areej Fatima; Tahir Abbas Khan; Taher M. Ghazal; Munir Ahmad; C. Yeun; Muhammad Attique Khanยท International Journal of Intelligent Systemsยท 2023ยท DOI 10.1155/2023/2599161

The core problem

The diagnosis of human diseases, especially deadly cancer diseases, remains a challenging task in the era of Healthcare 5.0. Accurate prediction is crucial for human wellbeing. The global Internet of Medical Things (IoMT) industry has evolved rapidly, but this advancement raises data privacy concerns. To ensure patient data privacy and fast data transmission, this work integrates federated deep extreme learning with edge computing for an intelligent lung disease diagnosis system. The proposed system employs federated deep extreme machine learning for lung disease prediction, enhanced by a fused weighted deep extreme machine learning methodology for improved accuracy. The model is validated using MATLAB 2020a, achieving 97.2% accuracy, surpassing state-of-the-art methods.

Innovation

The proposed intelligent system for lung cancer prediction combines federated learning, deep extreme learning machines (DELM), and edge computing. Federated learning ensures privacy by training models locally on distributed data without sharing raw patient data. Deep extreme learning machines provide fast learning and good generalization. A fused weighted approach is adopted to combine multiple DELM models, enhancing prediction performance. The system architecture is illustrated below:

The fused weighted method assigns weights to each local model based on performance, and the ensemble prediction is given by:

where are the weights, are the local model predictions, and is the number of models. The weights are optimized to minimize prediction error. The model is trained and tested on lung cancer datasets, with simulations conducted in MATLAB 2020a.

Introduction
The diagnosis of human diseases, especially deadly cancer diseases, remains a challenging task in the era of Healthcare 5.0. Accurate prediction is crucial for human wellbeing. The global Internet of Medical Things (IoMT) industry has evolved rapidly, but this advancement raises data privacy concerns. To ensure patient data privacy and fast data transmission, this work integrates federated deep extreme learning with edge computing for an intelligent lung disease diagnosis system. The proposed system employs federated deep extreme machine learning for lung disease prediction, enhanced by a fused weighted deep extreme machine learning methodology for improved accuracy. The model is validated using MATLAB 2020a, achieving 97.2% accuracy, surpassing state-of-the-art methods.

Why it matters

The integration of federated learning with deep extreme learning machines and fused weighting provides a robust solution for lung cancer prediction in Healthcare 5.0. Federated learning addresses privacy concerns by keeping data local, while edge computing ensures low-latency processing. The fused weighted ensemble enhances model accuracy by combining multiple local models, mitigating the limitations of individual models. The achieved accuracy of 97.2% demonstrates the model's superiority. However, challenges such as communication overhead and heterogeneity of edge devices need further investigation. Future work could explore scalability and real-world deployment.

Who should read this

Pembaca Research Digest dan praktisi rumpun ini.

Opening member contentโ€ฆ