Reinforcement learning for mobile robots

Deep Reinforcement Learning have proven robustness in many areas such as robotics, games. In Deep Reinforcement Learning, neural networks have strong ability to deal with high dimensional data, a good mean to learn features and functional approximation while reinforcement learning can make a system learn itself for a new goal and in a new environment.
From success of deep reinforcement learning, we are motivated to investigate methods that enable unmanned vehicle remember environment, identify location of target objects and navigate themselves to the target.
In the presentation, we briefly review notion of reinforcement learning and how deep learning enhance the success of reinforcement learning recent years. Some related research and toolboxes also will be introduced. Lastly, we will discuss challenges and some achievement in our research project.

Dr. Nguyen Do Van – Vietnam National University (Vietnam)

Dr. Nguyen Do Van received B.Eng. degree in Information Technology from Le Quy Don University in 2004 with honor, M.Eng and PhD degree in 2011 and 2014 respectively, in Information Science and Control Engineering from Nagaoka University of Technology, Japan.
He currently is a researcher at Institute of Information Technology, MIST. He also is an adjunct lecturer at Faculty of Information Technology, University of Engineering and Technology, Vietnam National University, Hanoi.
His research interest includes Machine Learning (Deep Learning, Reinforcement Learning), Data Mining (Statistical Model, Big Data), Intelligent Systems and Robotics. He is now advising an industrial project on Graph representation learning for social network mining and leading a research project on Deep reinforcement learning for intelligent mobile robots.
Dr. Van is member of the Institute of Electrical and Electronics Engineers (IEEE), International Rough Set Society (IRSS) and The Vietnamese Association for Pattern Recognition (VAPR). He was the recipient of Vietnam Intelligence Award in 2003 and AIAI Best Paper Adward in 2015.