An IoT-based, two-tier comprehensive and affordable intrusion detection system for smart home security
B. Suresh Ram, Terli Satyanarayana, U. Saritha, B. Sinduja · 2025
A smart home’s security system should be able to identify potential threats from network, gadgets, sensors, or users. In this research, we employ intrusion detection with machine learning algorithms like random forest (RF). Using a two-tiered method, in particular, we find anomalous actions that can take place in smart home (SH) environment. Using the dataset, we execute a number of machine learning classification models, including decision trees, xgboost, and random forests. Our tests reveal that while the algorithms are trained on decision layer, the models’ accuracy in spotting potential abnormalities that point to assaults varies. Our approach’s level of accuracy is encouraging for its potential application in SH system.