<p>Due to the advent of Mobile Edge Computing (MEC), Internet of Things (IoT) now has the opportunity to take advantage of the computing capabilities of the closest edge devices, which allows them to perform processing locally and not rely on centralized cloud servers. MEC also adds more processing and storage capabilities to network-edge devices in order to address the high delay requirements of mobile applications. But with the limited storage and computation capability of edge servers, the caching of only the most important application data is feasible, and hence best caching decisions are important in minimizing latency and lessening energy usage in edge settings. This study paper focuses on cooperative task offloading and data caching schemes that aim at decreasing the total latency on mobile computers. A combined approach of task offloading, data caching, and task prediction, which is enabled by a Bidirectional Long Short-Term Memory (Bi-LSTM) model, is presented to enhance the performance of Mobile Edge Computing (MEC) settings. The strategy uses adaptive task migration and smart caching techniques which are informed by the performance of incoming tasks and available edge node performance. A joint task migration, task offloading, and dynamic data caching of computational workloads and content is done using an Optimized Deep Q-Network (ODQN) algorithm. The framework uses Bi-LSTM-based prediction of tasks to make better edge scheduling, as well as offloading choices. The accuracy of prediction is 96.78% which is better than conventional models. The outcome of the simulation proves that the proposed model shows a significant advantage over traditional approaches to offloading tasks and caching data with a significant decrease in the average latency of services across mobile devices.</p>