Liang Ma  Liang Ma photo         

contact information

Research Staff Member - Distributed AI
Thomas J. Watson Research Center, Yorktown Heights, NY USA
  +1dash914dash945dash1247

links

Professional Associations

Professional Associations:  ACM  |  IEEE

profile


 

Liang Ma is a Research Staff Member at IBM T.J. Watson Research Center, Yorktown Heights, NY. Dr. Ma received the Ph.D. degree from the Department of Electrical and Electronic Engineering of Imperial College London in July, 2014. During his Ph.D. study, he was fully supported by the U.S.-U.K. ITA (International Technology Alliance) project, EPSRC (Engineering & Physical Sciences Research Council, U.K.), and departmental scholarships. He received both the M.Sc. and B.Sc. degree with distinction from the Beijing University of Posts and Telecommunications (BUPT), China. He was a member in the Lab of Electromagnetic Compatibility (EMC) and Wireless Communication Technologies at BUPT, participating National 863/973 Projects during 2007-2010. His current research is focusing on AI, Deep Neural Networks, Partial Observable Markov Decision Process (POMDP), Reinforcement Learning, Embedding, Question Answering, Graph Algorithms, Distributed Computing, and Cloud Computing.

Before joining Imperial College, he once worked as a research intern at NTT DoCoMo Beijing Labs, Ericsson (China) Communications, and Microsoft Research Asia (MSRA), where he was involved in WLAN Medium Access Control, High-speed Switching System, and Software Radio-Based Gigabit Multi-antenna Communications, respectively. Starting from 2011, he has been working with IBM T.J. Watson Research Center, Army Research Lab, and UMASS-Amherst on network analytics, machine learning, and algorithm design.

At IBM T.J. Watson Research Center, Dr. Ma is leading two projects with team members from Yale, UCSB, Northwestern, Penn State, Imperial College, U.S. NRL, U.S. ARL, and U.K. Dstl. In particular, one project is on hybrid networks supported by the International Technology Alliance (ITA). The project objective is to develop efficient resource management strategies via reinforcement learning in distributed SDN (software-defined networks), low-dimensional node sequence embedding to represent workflows, and accurate network performance inference using deep neural networks and LSTM. Another project that Dr. Ma is leading is on query answering in dynamic knowledge networks supported by the Network Science Collaborative Technology Alliance(NS-CTA). The project goal is to understand the question answering efficiency in dynamic and multi-genre networks, and employ such understanding to develop reliable query routing/answering schemes in socio-information networks. In addition, as the network may experience fragmentations due to high dynamicity, efficient algorithms such as social link maintenance and influence maximization are also explored in this project.

Dr. Ma has been an IEEE member since 2009, and an ACM member since 2013. He has served as a peer reviewer in a range of journals and conferences, including ACM TKDD, IEEE/ACM TON, INFOCOM, SECON, WCNC, etc. He was the recipient of IEEE International Conference on Communications (ICC 2019) Best Paper Award on reinforcement-learning approaches for resource management, IEEE International Conference on Distributed Computing System (ICDCS 2013) Best Paper Award, IBM Invention Patent Award, Chatschik Bisdikian Memorial Best Student Paper Award of ITA in Network & Information Sciences 2013, ACM SIGCOMM Internet Measurement Conference (IMC 2013) Best Paper Award Finalist, and the winner of Outstanding Graduate Student 2008 and Excellent Student Awards four times during 2003-2006.

 

Liang Ma's CV