The study focuses on comparative analysis of performances of accurate machine learning algorithms, Random Forest, and Grey Wolf Optimization (GWO) for determining lung cancer outcomes. The data analysis methodology followed a systematic process to enhance the quality of the dataset through preprocessing which included data cleaning, balancing of classes using Synthetic Minority Oversampling Technique (SMOTE), and dealing with outlying values which was done by using Interquartile Range (IQR). Using the GWO algorithm for the feature selection process allowed the suggestion of critical health-related indicators, which impacted the model’s performance. In terms of performance, the Random Forest classifier performed outstandingly by having an accuracy of 97.21% along with precision and recall of 95.02% & 92.11% respectively. These statistics support evidence regarding the type of lung cancer the model can accurately predict, which is important to clinical diagnosis. The results are consistent with extant literature while providing new understanding concerning the efficacy of the proposed machine learning algorithms in healthcare. This study shows how lung cancer prognosis might be enhanced by advanced analytics for diagnosis and subsequent treatment, opening up a wide range of future research to support its application, verify the outcomes, and increase diagnostic efficiency in oncological practice.