With the increasing competition for advertisement space on Facebook, the prevalence of automated bidding by bots poses a challenge for legitimate businesses. This paper explores the detection of anomalous bot-users in the context of Facebook Ads bidding, leveraging data from the Kaggle competition “Facebook Recruiting IV: Human or Robot?” Our approach involves comparing various anomaly detection methods, including group-based, temporal-based, and edge-attributed graph-based approaches. I implement algorithms such as BIRDNEST, Support Vector Machines, Logistic Regression, Adaboost, Gradient Boosting, Random Forests, Bagging, and Extra Trees. The results, evaluated using the area under the ROC curve (AUC), highlight the effectiveness of Gradient Boosting and other boosting methods in achieving competitive performance. The study also discusses insights gained from feature engineering and the challenges of scalability in group-based and graph-based approaches. Overall, the research contributes to the ongoing efforts to combat auction sniping and enhance the integrity of Facebook’s advertisement space.