Predicting and Classifying Lung Cancer Using KNN Classification Machine Learning Algorithm

Lung cancer remains one of the most prevalent and deadly forms of cancer globally, necessitating early detection and accurate prediction to improve patient outcomes. This paper presents the development of a machine learning-based model to predict lung cancer levels—low, medium, or high—using a dataset containing various clinical and lifestyle attributes such as age, gender, air pollution exposure, smoking habits, and genetic risk. Multiple machine learning algorithms, including Logistic Regression (LR), Decision Trees (DT), Random Forest (RF), K-Nearest Neighbors (KNN), and others, were applied to the dataset to assess their prediction performance. The results revealed that Random Forest (RF) yielded the highest accuracy of 100%, precision of 98%, sensitivity of 97%, and an F1-score of 95%, making it the most effective model for lung cancer prediction. Decision Trees (DT) followed with an accuracy of 88%, while K-Nearest Neighbors (KNN) and Logistic Regression (LR) achieved accuracies of 87% and 81%, respectively. Kernel Density Estimate (KDE) plots were utilized to assess the distribution of features, providing insights into feature selection and class separation. The study highlights the significance of accurate feature selection and the efficacy of machine learning models in predicting lung cancer severity, paving the way for better diagnostic tools in medical practice.