Rikesh Yadav
Investigating sample-efficient Reinforcement Learning and Graph Neural Networks for trustworthy decision-making.
How can autonomous agents learn optimal, resilient strategies in dynamic networks? My research solves distributed decision-making bottlenecks for Edge Intelligence.
Currently architecting adaptive evacuation frameworks and decentralized edge-caching models as a Research Intern at the INEC Laboratory, Taiwan. The current methodology integrates Graph Convolutional Networks with Deep Q-Learning to optimize multi-agent resource allocation under strict latency constraints.
The immediate objective is to translate theoretical reinforcement learning paradigms into real-time, executable algorithms. My long-term vision is to develop scalable, trustworthy AI systems capable of zero-shot generalization in mission-critical environments.
.png&w=3840&q=75)
About Me
Research Intern
Intelligent Networks & Edge-Cloud Computing (INEC) Laboratory
Yuan Ze University, Taiwan
My interest in research stems from a deep curiosity about how intelligent systems can learn, adapt, and reason within complex environments. I am driven by the challenge of moving beyond static algorithms to develop dynamic, learning-based architectures that can handle uncertainty and make autonomous decisions.
I approach learning with a disciplined mindset, balancing theoretical exploration with rigorous experimentation. Whether dissecting the mathematical foundations of a new paper or engineering a robust training environment, I believe that true innovation happens at the intersection of critical thinking and practical, hands-on implementation.
Ultimately, I aspire to become a researcher who bridges the gap between advanced theoretical concepts and scalable real-world solutions. Through continuous learning and a commitment to trustworthy AI, I hope to shape intelligent systems that are not only highly capable, but also interpretable, safe, and deeply impactful.
"I believe impactful artificial intelligence should not only achieve high performance but also make reliable, interpretable, and adaptive decisions in complex real-world environments. My goal is to contribute to intelligent systems that combine strong theoretical foundations with meaningful practical impact."
Research Interests
Reinforcement Learning
Designing intelligent agents capable of learning optimal behaviour through interaction with dynamic environments.
Deep Reinforcement Learning
Combining deep neural networks with reinforcement learning to solve large-scale sequential decision-making problems.
Graph Neural Networks
Learning meaningful representations from graph-structured data for intelligent reasoning and relational learning.
Edge Computing
Developing distributed intelligent systems capable of efficient computation across edge-cloud environments.
Trustworthy AI
Building reliable, interpretable, and robust AI systems that can be deployed responsibly in real-world applications.
Human-Centered AI
Designing intelligent systems that augment human decision-making while prioritizing usability, transparency, and societal benefit.
Research Philosophy
"I am interested in building intelligent systems that can perceive, reason, and adapt to complex real-world environments. My long-term goal is to contribute to trustworthy and scalable AI through research at the intersection of reinforcement learning, graph learning, and intelligent systems."
Current Research
Adaptive Evacuation: A GNN-Enhanced Deep Reinforcement Learning Framework
Developing a hybrid framework combining Graph Convolutional Networks (GCNs) with Deep Q-Networks (DQNs) to improve adaptive decision-making in IoT-enabled indoor environments.
Explore ResearchResearch Output
Adaptive Evacuation: A GNN-Enhanced Deep Reinforcement Learning Framework for Intelligent Indoor Emergency Response
Developing a hybrid Graph Neural Network and Deep Reinforcement Learning framework for adaptive indoor emergency response in IoT-enabled environments.
A Novel Adaptive Caching Framework for Dynamic Web Portals Using Machine Learning-Aided Predictive Models
Explored adaptive caching strategies using machine learning to improve dynamic web portal performance.
Research Projects
Adaptive Evacuation
Developing a Graph Neural Network enhanced Deep Reinforcement Learning framework for intelligent indoor emergency evacuation under dynamic emergency conditions.
MindMate
AI-assisted mental wellness platform integrating conversational AI, journaling, mood tracking, and personalized recommendations.
Latest Entries
Let's Connect
I am always interested in discussing research ideas, graduate opportunities, collaborations, and intelligent systems. Whether you are a researcher, professor, student, or industry professional, I would be happy to connect.
Coimbatore, India
Currently conducting research in Taiwan.
Open to Collaboration
"I welcome discussions on research collaborations, graduate opportunities, intelligent systems, reinforcement learning, graph neural networks, and related areas of artificial intelligence."
"The best ideas often begin with meaningful conversations. I look forward to connecting with researchers and collaborators who share an interest in advancing intelligent systems."