The Internet of Things (IoT) is becoming a necessary part of everyday life as billions of devices are now online. IoT devices are linked together making them vulnerable to cyber intrusions. Cyber intrusions on IoT networks have increased significantly in terms of both number and effectiveness in the last few years. A novel approach for predicting cyber-attacks on IoTs devices is presented in this research. The enhanced Optimal Weighted Model for Decision Making (OWM-DM) technique is employed in the proposed framework. The proposed OWM-DM is used to classify IoT signals and identify intrusions. The OWM-DM classifies the features most important for detecting attacks and is then used to optimize the model. An IoT traffic dataset from the actual world was used to assess the suggested architecture. The outcomes demonstrate that the architecture can accurately classify threats. To improve the accuracy of the approach, the architecture may identify which features are most important for intrusion classification. The suggested method concentrates on internet traffic features that may indicate cyber-security risks to IoT networks and impacted nodes in the network. Specific feature vectors have been generated using the components obtained on each IoT console. The suggested OWM-DM can considerably anticipate the multiple stages of cyber violence with 99% accuracy, according to the evaluation findings. Handling cyber-security attacks in real-time communication requires such precise forecasting. Due to its efficacy and flexibility, the framework can be utilized to forecast intrusions in actual time, even in large-scale IoT deployments. Its ability to compute efficiently enables it to produce precise predictions quickly, enabling timely detection and intervention. By identifying and mitigating potential points of entry for attacks, the framework contributes to enhancing the security of IoT networks.