<![CDATA[Lung cancer is one of the leading causes of global mortality, often difficult to detect early due to its nonspecific initial symptoms. This study proposes a Machine Learning-based approach to classify lung cancer risks using the Random Forest algorithm optimized with GridSearchCV. The identified research gap is the lack of interactive web-based implementations that deliver real-time classification results with a user-friendly interface for general users. The objective of this study is to develop an accurate and efficient classification model and integrate it into a web application using Streamlit. The dataset was sourced from Kaggle, consisting of 5,000 patient records and 18 clinical and lifestyle-related features. The preprocessing steps included data cleaning, normalization, encoding, and feature recategorization. Model performance evaluation using Accuracy, Precision, Recall, and F1-Score metrics showed an accuracy of 90%. Feature importance analysis identified smoking habits, throat discomfort, and respiratory issues as dominant predictors of lung cancer. The model was then deployed into a Streamlit-based web application and tested via a User Acceptance Test (UAT) involving 50 respondents, resulting in a Mean Opinion Score (MOS) average above 84%. These findings indicate that the developed prediction system is not only technically accurate but also well-accepted by users, highlighting its potential as a practical and efficient tool for early lung cancer screening.]]>
Copyrights © 2025
