Mehran Taghian

About

Mehran Taghian is at the forefront of making intelligent systems both more transparent and more resilient. His research focuses on the critical intersection of **explainable artificial intelligence (XAI)** and **reinforcement learning (RL)** , with a particular emphasis on **robotics** and **industrial automation**. In his highly cited 2024 work, Taghian pioneered the application of **Layer-wise Relevance Propagation (LRP)** to demystify the decision-making processes of deep reinforcement learning agents in robotic domains. This breakthrough, garnering 9 citations, directly addresses a major criticism of DRL by providing a clear, visual map of why an agent chooses a specific action, which is essential for trust and safety in autonomous systems. Complementing this, his research on enhancing **hardware fault tolerance** demonstrates a practical application of RL policy gradient algorithms. Here, Taghian moves beyond traditional, costly hardware duplication by developing adaptive algorithms that allow machines to autonomously detect and reconfigure around faults. By bridging the gap between theoretical AI explainability and real-world industrial robustness, Taghian’s work is laying the essential groundwork for the next generation of reliable, transparent, and truly autonomous machines.

Research Focus

Key Achievements

Top Papers

1. 1
   Explainability of deep reinforcement learning algorithms in robotic domains by using Layer-wise Relevance Propagation

   9 citations · 2024
2. 2
   Enhancing Hardware Fault Tolerance in Machines with Reinforcement Learning Policy Gradient Algorithms

   2 citations · 2024

Key Collaborators

Contact & Links