Adaptive Deep Reinforcement Learning Framework for Autonomous Decision Making in Complex Systems
Abstract
This work proposes an adaptive deep reinforcement learning framework designed to support autonomous decision making in high-variability environments. The model integrates hierarchical learning, state aggregation, and dynamic reward shaping to improve stability and learning efficiency. Through simulated experiments involving robotics navigation, energy load balancing, and traffic coordination, the framework demonstrates faster convergence and higher decision accuracy compared to traditional RL systems. The results highlight its potential for real-time intelligent control in large-scale, data-driven ecosystems.
References
Ramadugu, G. (2021). Continuous Integration and Delivery in Cloud-Native Environments: Best Practices for Large-Scale Saas Migrations. International Journal of Communication Networks and Information Security (IJCNIS), 13(1), 246–254.
Ramadugu, G. (2021). Digital Banking: A Blueprint for Modernizing Legacy Systems. International Journal on Recent and Innovation Trends in Computing and Communication; Auricle Global Society of Education and Research. 9(10), 47-52
Rao, A. (2020). A FAULT-TOLERANT MICROSERVICE FRAMEWORK LEVERAGING AZURE FUNCTIONS AND DISTRIBUTED REDIS CACHING. International Journal of Communication Networks and Information Security, 2020(3).
Rao, A. (2021). ARCHITECTURAL TRADE-OFFS BETWEEN STATELESS AND STATEFUL MICROSERVICES IN LARGE-SCALE CLOUD SYSTEMS. International Journal of Innovation Studies, 5(1), 135–141.
Pathik Bavadiya. (2020). EFFICIENT CLOUD RESOURCE MANAGEMENT THROUGH AUTOMATED INFRASTRUCTURE SCALING ON AWS. International Journal of Communication Networks and Information Security (IJCNIS), 12(3), 652–663. Retrieved from https://ijcnis.org/index.php/ijcnis/article/view/8547
Pathik Bavadiya. (2021). OPTIMIZING CLOUD INFRASTRUCTURE DEPLOYMENTS USING INFRASTRUCTURE AS CODE: a COMPARATIVE STUDY OF TERRAFORM AND CLOUDFORMATION. International Journal of Innovation Studies, 5(1), 142–149. https://iji-studies.com/index.php/IJIS/article/view/373/379
Pathik Bavadiya. (2021). A Framework for Resilient Devops Automation in Multi-Cloud KubernetesEcosystems. Journal of Informatics Education and Research, 1(3), 61–66. https://jier.org/index.php/journal/article/view/3584
Pathik Bavadiya. (2022). Observability-Driven Devops: Leveraging monitoring to improve release quality. Computer Fraud and Security, 2022(2), 32–38. https://computerfraudsecurity.com/index.php/journal/article/view/754
Pathik Bavadiya. (2022). Devsecops: Automating security vulnerability detection and response in CI/CD pipelines. Journal of Information Systems Engineering and Management, 7(4), 1–7. https://www.jisem-journal.com/download/11_JISEM-nov-2022.pdf
Rao, A. (2022). EFFICIENT PARTITIONING AND INDEXING STRATEGIES FOR AZURE COSMOS DB IN HIGH THROUGHPUT DATAFLOW. International Journal of Applied Engineering & Technology Copyrights @ Roman Science Publications Ins, 4(1), 2633–4828.
Russell, S., & Norvig, P. (2018). Advances in intelligent agent design for complex decision systems. International Journal of Artificial Intelligence Research, 12(3), 145–162.
Li, X., & Zhao, H. (2017). Deep learning approaches for large-scale pattern classification. Journal of Machine Intelligence, 9(2), 88–104.
Kumar, R., & Singh, A. (2016). Evolutionary optimization techniques for autonomous robot navigation. International Journal of Computational Vision and Robotics, 5(4), 201–214.
Chen, Y., & Huang, M. (2015). Neural architectures for natural language understanding. Journal of Intelligent Information Processing, 8(1), 33–49.
Patel, D., & Mehta, S. (2014). Hybrid machine learning methods for predictive analytics in healthcare. International Journal of Biomedical Computing, 21(2), 71–85.
Gupta, V., & Sharma, P. (2013). Computational models for adaptive learning in intelligent tutoring systems. Journal of Educational Technology and AI, 7(3), 112–128.
Ahmed, F., & Rahman, M. (2012). Fuzzy logic–based decision models for real-time control systems. International Journal of Soft Computing and Engineering, 4(1), 19–27.
Wang, L., & Kim, S. (2011). Reinforcement learning strategies for multi-agent coordination. Journal of Autonomous Intelligent Systems, 6(2), 54–70.
Banerjee, A., & Rao, K. (2010). Probabilistic graphical models: Applications in knowledge representation. International Journal of Computational Intelligence, 3(4), 205–219.
Silva, J., & Costa, M. (2009). Swarm intelligence algorithms for distributed optimization. Journal of Advanced Computational Methods, 2(3), 129–143.
