A Self – Adaptive Reinforcement Learning Driven Teaching Management System Based on Cloud Computing Technology
G. Vedha Sree1, G. Chinna Karthik2, K. Divya Sri3, K. Ravi Kishore4, J. Unni Kiran5
1Information Technology & Prasad V. Potluri Siddhartha Institute of Technology (PVPSIT)
2Information Technology & Prasad V. Potluri Siddhartha Institute of Technology (PVPSIT)
3Information Technology & Prasad V. Potluri Siddhartha Institute of Technology (PVPSIT)
4Information Technology & Prasad V. Potluri Siddhartha Institute of Technology (PVPSIT)
5Information Technology & Prasad V. Potluri Siddhartha Institute of Technology (PVPSIT)
Abstract - The introduction of cloud computing has greatly enhanced the scale of and access to today's modern teaching management systems. However, current e-learning platforms that are hosted in the cloud typically use static resource allocation strategies, which lead to poor resource usage and limited adaptability for fluctuating workloads [1]–[5]. To solve these problems, this paper introduces a new type of teaching management system with self-adaptive resource allocation using Deep Reinforcement Learning (DRL) for cloud resource management. The proposed self-adaptive system will allocate cloud resources using real-time workload data and also continually improve/drill down to the optimal resource allocation policy, improving performance and lowering operational costs [11]–[14]. The proposed system is envisioned to integrate both cloud-based education services as well as adaptive education strategies, thus enhancing scalability, responsiveness, and overall system efficiency. Based on the results from preliminary experiments, the proposed system is shown to provide a more intelligent and adaptive solution than other standard cloud-based teaching management systems.
Key Words: Cloud Computing, Teaching Management System, E-learning, Deep Reinforcement Learning, Resource Allocation, Adaptive Systems, Learning Analytics, Cloud Resource Management, Educational Technology, Intelligent Systems
