Research Interests

My research is in the general areas of machine learning and artificial intelligence. I specialize on clinical data mining, biomedical predictions, data analysis for healthcare and wearable sensors, planning, and optimization. I am interested in both fundamental research and real-world applications. More specifically, I have performed research on:

  • Machine learning (representation, interpretability, efficiency, compactness, clinical applications)
  • Clinical and healthcare data mining (real-time data analysis and decision support for ICU/OR/general wards/wearables)
  • Scalable optimization and learning algorithms
  • Nonlinear optimization (networking, biological systems, radiotherapy)
  • Automated planning (supply chain, transportation, airlines)

I have also seriously studied several other areas through inter-disciplinary collaboration, including cyber-physical systems, wireless sensor networks, clinical data analysis, radiation oncology, and metabolic networks.

Research Synopsis

  • Machine learning and data mining. Our goal in this area is to develop efficient and effective algorithms to extract knowledge from massive amount of data. In particular, we focus on learning stream data (temporally ordered real-time data flow) and sequential data. These two types of data have extremely wide applications, from business transactions to bioinformatics, and are more challenging to analyze than static data. Our recent research contributions include:
    • Deep learning   (KDD-16, arXiv-16, ICML-15)
    • Scalable and interpretable machine learning   (DMKD-02, VLDB-02, DPD-05, TKDE-06, TKDE-09, KAIS-11, KDD-13, KDD-14, KDD-15, Aistats-15, AAAI-15)
    • Mining and learning on real-time data and stream data (KDD-07, AAAI-08, KDD-09, TKDD-09, AAAI-14)
    • Dimensionality reduction and manifold learning (ICTAI-09, ICML-11, AI-11, ICML-13, AAAI-13, ICDM-13, TKDE-15)
    • Data mining and learning for medical and clinical data (ICMLA-09, Neuro-09, AIMA-11, BioDM-11, ARO-12, IJKDB-12, KDD-12, JHM-13, IPSN-14, AMIA-15)
    • Machine learning for Web queries (TKDE-11, CIKM-11)
  • Computational optimization. Optimization is a computational problem with broad scientific and engineering applications. We have been working on the theory and algorithms for improving the quality and time cost of optimization. We are also studying applications of optimization. Our recent research contributions include:
    • Decomposition-based large-scale optimization (AIJ-05, CP-05, PAAP-08)
    • Nonlinear optimization theory (JOGO-07, COA-08, GrC-08)
    • Intensity modulated radiotherapy optimization (LAA-07, ICCR-07)
    • Wireless sensor network optimization (RTSS-08, MobiHoc-10, RTSS-10, ICDCS-11, ECRTS-11, RTAS-11, TPDS-11, RTAS-12, InfoCom-12, TPDS-12, TECS-13, IPSN-14, IWQoS-14, RTSS-15, IoTDI-16)
    • Design optimization for streaming applications (SAAHPC-10, ASAP-11, ICPADS-12,PPoPP-13)
    • Nonlinear optimization in biological systems (PLoS-12, BMC-12)
  • Planning and search. Planning entails automatically finding a course of actions to accomplish goals under logical and resource constraints. I investigate the fundamental search algorithms for planning as well as practical aspects of planning systems. Our recent research contributions include:
    • Partial order based search space reduction algorithms (IJCAI-09a, IJCAI-09b, arXiv-11)
    • Efficient satisficing planning (IJCAI-07a, AAAI-08)
    • Cloud computing and planning (CCV-10, CloudCom-12)
    • Long distance mutual exclusion (IJCAI-07b, AIJ-09)
    • Optimal planning and search (ICAPS-06, CPAIOR-06, AAAI-10, JAIR-12, AAAI-13)
    • Optimal temporally expressive planning (ICAPS-09, TIST-12)
    • Planning applications (mobile computing, Web services) (SPARK-11, AAAI-12, TSC-12)

Related Information