In the burgeoning era of Mobile Edge Computing (MEC), efficient resource scheduling stands paramount to ensuring seamless application performance and user experience. While existing scheduling models offer a foundation for task-to-VM mapping, they often suffer from a lack of adaptability to dynamic environmental changes, inability to anticipate future tasks, and sub-optimal resource utilization, leading to increased makespan and missed deadlines. Addressing these challenges, this paper introduces a novel design for an Incremental Learning Model aimed at enhancing the resource scheduling efficiency of mobile edge deployments. At its core, the model employs temporal task metrics such as dependency, makespan, deadline, RAM, bandwidth, frequency, and requesting IP for a meticulous task analysis. While, edge VMs are scrutinized using metrics such as RAM, Bandwidth, Processing Elements, MIPS, distance to requesting IP, and frequency of use. To bridge the gap between tasks and VMs, we utilize the Grey Wolf Firefly (GWFF) Optimizer, renowned for its precision and speed. A standout feature of our approach is the integration of the VARMAx-based model into the GWFF Optimizer. This combination not only assists in anticipating future tasks but also dynamically adjusts internal VM metrics. The result is a scheduler that’s not just reactive but also proactive in nature. Our evaluations have showcased remarkable improvements in performance metrics. Compared to recent scheduling models, we have achieved a 3.9% reduction in makespan, 4.5% enhancement in deadline hit ratio, 2.9% boost in scheduling efficiency, and a commendable 2.4% decrease in delay. In conclusion, this work signifies a pivotal stride forward in the realm of MEC scheduling, paving the way for more responsive, efficient, and intelligent mobile edge deployments in the future. The proposed model not only addresses the limitations of contemporary systems but also lays the groundwork for emerging research to build upon, thus having a profound impact on the evolution of edge computing paradigms.