Analysis of Lung Cancer for Developing Smart Healthcare with the Help of BGWO Based TSA-XGBoost Model

Given the circumstances the healthcare system, the Internet of Things (IoT) is crucial. IoT gadgets offer patient data for the framework of healthcare monitoring. IoT is a key component in every aspect of the health care management system since people can use smart devices to check on their health....

Full description

Saved in:

Bibliographic Details
Published in:	2023 International Conference on Self Sustainable Artificial Intelligence Systems (ICSSAS) pp. 910 - 916
Main Authors:	R, Mahaveerakannan, S, Murali Dhar M, Goswami, Sohan, Krishna, P. Gopala, Jayanthi, K., Kumar, M. Sathish
Format:	Conference Proceeding
Language:	English
Published:	IEEE 18.10.2023
Subjects:	Analytical models Binary grey wolf optimization algorithm Classification algorithms Extreme gradient boosting Feature extraction Feature selection Internet of Things Lung cancer Machine learning algorithms Medical services Synthetic minority oversampling methods
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Given the circumstances the healthcare system, the Internet of Things (IoT) is crucial. IoT gadgets offer patient data for the framework of healthcare monitoring. IoT is a key component in every aspect of the health care management system since people can use smart devices to check on their health. Lung cancer is a fatal malignancy, and the likelihood of survival is increased by early identification. Because of the computational difficulty involved in gathering characteristics, it is imperative to design an approach using machine learning techniques for categorising cancer disease because the classification results given by the current methods are inadequate. With the use of the synthetic minority oversampling methods (SMOTE) methodology, the incoming data is pre-processed and balanced. With the help of the binary grey wolf optimisation algorithm (BGWOA), the pertinent features are best chosen. Finally, the suggested model's hyper-parameters are best chosen by the tunicate swarm optimisation (TSA) model, and the classification is carried out by the extreme gradient boosting (XGBoost) model. The experimental analysis demonstrates that the suggested model attained accuracy and recall values of 98% and 95%, respectively, compared to 95% and 95%, respectively, for the identical proposed model without the feature selection (FS) method.
DOI:	10.1109/ICSSAS57918.2023.10331761