In-home security, the ubiquitous utilization of surveillance systems, has become increasingly common. This research delves into anomaly detection within CCTV footage, specifically focusing on identifying and analyzing infrequent events indicative of potential security threats to residential space. This study conducts a comparative analysis of major machine-learning approaches for infrequent itemset mining. The research employs a comprehensive data set of real-world CCTV footage, capturing various scenarios encountered in home environments. Performance metrics, including accuracy, recall, and F1-score, are systematically evaluated to discern the strengths and limitations of each algorithm. By shedding light on rare patterns, this research equips homeowners, security practitioners, and system developers with valuable insights for selecting and implementing anomaly detection solutions tailored to safeguard home premises. The comparative nature of the study offers a nuanced understanding of algorithmic capabilities, fostering informed decision-making and serving as a foundation for future advancements in residential security.