Lung cancer often affects smokers, starts in the lungs and manifests itself with a variety of signs and symptoms. Lung cancer can be predicted using machine learning algorithms, especially when used by descriptive intelligence (XAI). This study combines machine learning techniques with XAI methods to propose a robust model for future lung cancer risk assessment. Similar examination showed that different order techniques, for example, lo-gistic relapse, k-closest neighbors (KNN), support vector machine (SVM), and choice tree model accomplished 95% precision. Additionally, the prediction accuracy of lung cancer can be greatly improved using methods such as normalization (also known as Z-score normalization) and joint boosting algorithms (especially CAT Boost). Both the Random Forest Classifier and CAT Boost showed an accuracy of 98.38%, mostly due to normalization, a scaling process that increases search and improves the performance of some machine learning algorithms. Integrated descriptive intelligence (XAI) methods, such as LIME and SHAP values, can improve model interpretation by incorporating detailed information into specific predictions, understanding and teaching what is important. This analysis will impact the field of lung disease prediction by combining machine learning with XAI methods to facilitate the development of predictive tools and improvements in the clinical assessment of lung cancer-related factors. Ultimately, the aim of this research is to make early prediction of lung cancer and ultimately save lives.