Liping Tang

Papers

Total Citations

H-Index

About

Liping Tang is a researcher whose work bridges neural network theory and reinforcement learning, with a particular focus on overcoming the "curse of dimensionality" in continuous state-space systems. Their most cited contribution, the 2010 paper "Study on Q-learning algorithm based on ART2," introduces a novel integration of ART2 neural networks into the Q-learning framework. By leveraging ART2's ability to dynamically cluster and recognize patterns in real-time, Tang proposed a method that reduces the computational explosion typically faced when applying Q-learning to complex, high-dimensional environments. This work provides a clear, step-by-step algorithmic approach, offering a practical solution for intelligent systems requiring adaptive decision-making without predefined state discretization. While the paper has garnered 3 citations, its conceptual contribution lies in demonstrating how unsupervised neural architectures can enhance model-free reinforcement learning. Tang’s research is particularly relevant for students and engineers working on autonomous systems, robotics, or adaptive control, where efficient state representation is critical. The work stands as a thoughtful intersection of cognitive modeling and machine learning, highlighting Tang’s interest in making reinforcement learning more scalable and biologically inspired.