Reinforcement Learning Algorithms for Intelligence Enhancement and Continual Learning

Dear Colleagues,

This Special Issue, “Reinforcement Learning Algorithms for Intelligence Enhancement and Continual Learning”, aims to showcase high-quality research on the design, analysis, and application of reinforcement learning algorithms for developing lifelong learning and autonomous systems that can adapt to emerging intelligence enhancement. It is observed that many trained models are undergoing data shifting within their domains due to such developments. Likewise, problems with dynamic nature, or those subject to time effects need to adapt and encompass these new conditions. These include algorithmic innovations that enable learning agents to continuously acquire, retain, and transfer knowledge in dynamic environments.

The scope of this Special Issue encompasses, but is not limited to, the development and evaluation of reinforcement learning methods for continual and lifelong learning, including strategies for mitigating catastrophic forgetting, knowledge consolidation, and transfer and curriculum learning. Contributions addressing adaptive policy learning, meta-reinforcement learning, and self-improving agents are particularly encouraged.

At the system and application levels, this Special Issue invites research on reinforcement learning algorithms for intelligence enhancement in autonomous systems, robotics, smart environments, cyber–physical systems, and decision-support platforms. Submissions exploring multi-agent reinforcement learning, cooperative and competitive learning frameworks, and human-in-the-loop learning systems are also within scope.

Furthermore, this Special Issue welcomes studies on the integration of reinforcement learning with complementary computational approaches, such as deep learning, evolutionary algorithms, symbolic reasoning, and optimization techniques, to enhance robustness, scalability, and interpretability. Interdisciplinarity contributions on theoretical analysis, convergence guarantees, performance evaluation, and benchmarking in continual learning settings are highly encouraged.

In addition, the Special Issue invites submissions on novel algorithmic frameworks, hybrid learning architectures, and efficient training methodologies, including resource-aware learning, online learning, and real-time adaptation. All submissions should demonstrate clear methodological rigor and, where appropriate, validation through empirical studies, simulations, or real-world applications.

Topics include (but are not limited to):

Reinforcement learning for continual and lifelong learning;

Algorithms for mitigating catastrophic forgetting;

Meta-learning and adaptive reinforcement learning;

Robustness, privacy, and security of experience replay mechanisms in continual RL;

Reinforcement Learning strategies for adapting Vision Transformers (ViTs) and Large Language Models (LLMs);

Experience generalisation, transfer learning, and knowledge reuse;

Multi-agent and distributed reinforcement learning;

Agentic and autonomous learning systems;

Human-in-the-loop, active, and interactive learning;

Hybridization of RL with evolutionary and swarm intelligence methods;

Resource-efficient and scalable RL algorithms;

Evaluation frameworks and benchmarks for continual RL;

Explainable and interpretable reinforcement learning;

Robotics and smart systems;

Cyber–physical and cyber security systems;

Decision-support systems ;

Interdisciplinary problem solving.

Finally, I would like to thank Dr. Janaki Sivakumar and Dr. Engin Esme for their valuable assistance with this Special Issue.

Dr. Mehmet Aydin
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Algorithms is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• reinforcement learning
• continual learning
• catastrophic forgetting
• transfer learning
• meta-reinforcement learning
• multi-agent systems
• autonomous systems
• explainable AI