Xin Liu:
Autonomous Greenhouse Control: A Bayesian Optimization and Model-based Reinforcement Learning Approach (Invited Talk)

Xin Liu received her Ph.D. degree in electrical engineering from Purdue University in 2002. She is currently a Professor in Computer Science at the University of California, Davis. Her current research interests fall in the general areas of machine learning algorithm development and machine learning applications in human and animal healthcare, food systems, and communication networks. Her research on networking includes cellular networks, cognitive radio networks, wireless sensor networks, network information theory, network security, and IoT systems. She is an IEEE Fellow.

In this talk, we present our work on using machine learning techniques for autonomous greenhouse control. We participated in an autonomous greenhouse challenge. In this challenge, we were given limited access to a greenhouse simulator to grow lettuce. In this simulator, we can control the greenhouse operation, from screen material and plant spacing to indoor heating, illumination, and CO2 level, with the objective to maximize the net profit of the growing season. To achieve this goal, we design a machine learning approach based on Bayesian optimization and model-based reinforcement learning and achieve promising results.