AI-Driven Lung Cancer Screening: A Comparative Analysis of Machine Learning Models

Peilin Li, Shirui Lyu, Peter Niu

2024

Abstract

This study explores the application of artificial intelligence (AI) in the field of lung cancer screening, evaluating the performance of three machine learning models: Random Forest, K-Nearest Neighbors (KNN), and Decision Tree. The Random Forest model emerged as the most accurate, with an overall accuracy of 88.2% and a balanced performance across both classes, indicating its superior generalization capability for new data subsets. Feature importance analysis revealed that 'Age' was a significant predictor in both Random Forest and Decision Tree models, highlighting its predictive value in the dataset. The KNN model, while achieving an accuracy of 81.6%, exhibited a performance imbalance, particularly struggling with class 0 samples, likely due to insufficient clustering or separation between classes. The Decision Tree model's lower accuracy was attributed to potential overfitting in the training subset, capturing noise specific to the training data and reducing its generalization ability. Notably, 'Chronic Disease' was found to be a highly important feature in the Decision Tree model, suggesting a biased decision rule. Overall, the findings underscore the potential of AI in enhancing lung cancer screening and the importance of feature selection and model generalization in achieving accurate predictions.

Paper Citation

in Harvard Style

Li P., Lyu S. and Niu P. (2024). AI-Driven Lung Cancer Screening: A Comparative Analysis of Machine Learning Models. In Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM; ISBN 978-989-758-738-2, SciTePress, pages 310-314. DOI: 10.5220/0013329700004558

in Bibtex Style

@conference{mlscm24,
author={Peilin Li and Shirui Lyu and Peter Niu},
title={AI-Driven Lung Cancer Screening: A Comparative Analysis of Machine Learning Models},
booktitle={Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM},
year={2024},
pages={310-314},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013329700004558},
isbn={978-989-758-738-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM
TI - AI-Driven Lung Cancer Screening: A Comparative Analysis of Machine Learning Models
SN - 978-989-758-738-2
AU - Li P.
AU - Lyu S.
AU - Niu P.
PY - 2024
SP - 310
EP - 314
DO - 10.5220/0013329700004558
PB - SciTePress