The study titled “An Experimental Study of Binary Classification on Imbalanced Datasets” evaluates multiple classifiers—such as LGBMClassifier, Isolation Forest, XGBoost, and others—for their effectiveness in addressing class imbalance. Among these, LGBMClassifier emerges as the top performer, showcasing remarkable capability through the use of weighted loss functions and boosting techniques. This classifier excels in achieving high accuracy in binary classification tasks on datasets with uneven class distributions. The findings highlight LGBMClassifier’s robustness in handling challenging data scenarios, underscoring its practical relevance in real-world applications where class imbalance is prevalent. By demonstrating superior performance compared to other models tested, the study provides valuable guidance to practitioners seeking optimal solutions for binary classification on imbalanced datasets. This research contributes insights that enhance understanding of classifier selection strategies tailored to address class imbalance effectively.