Lung cancer is one of the top causes of death worldwide, emphasizing the crucial importance of reliable and effective detection technologies for timely diagnosis and treatment. In this article, we provide a machine learning-based method for detecting lung cancer using a dataset with binary attributes (Yes/No). The methodology makes use of the Light Gradient Boosting Machine (LightGBM) classifier, which is well-suited to handle structured binary data and capturing complicated relationships within the feature space. The dataset is rigorously pre-processed, including binary encoding of categorical variables and gender representation, to ensure compliance with the classifiers. Following a methodical train-test split, the LightGBM model is trained and tested using performance metrics like accuracy, precision, recall, F1-score, and AUC-ROC. The results show that both classifiers attain good prediction accuracy, with LightGBM outperforming in computational efficiency. The proposed methodology shows the potential of merging modern machine learning and deep learning techniques to identify lung cancer. This method not only improves detection accuracy but also offers significant insights to healthcare practitioners.