Physics

Action (physics)

Mathematics

Reinforcement learning

Markov chain

Statistics

Markov decision process

Mathematics education

Machine learning

Artificial intelligence

Markov process

Action learning

Computer science

Cooperative learning

Quantum mechanics

Teaching method

IEEE/CAA Journal of Automatica Sinica

In this paper, a new machine learning framework is developed for complex system control, called parallel reinforcement learning. To overcome data deficiency of current data-driven algorithms, a parallel system is built to improve complex learning system by self-guidance. Based on the Markov chain (MC) theory, we combine the transfer learning, predictive learning, deep learning and reinforcement learning to tackle the data and action processes and to express the knowledge. Parallel reinforcement learning framework is formulated and several case studies for real-world problems are finally introduced.

Parallel reinforcement learning: a framework and case study