Research

The Virginia Tech Machine Learning Laboratory research focuses on the development of fundamental machine learning (ML) techniques that address the problems of robust ML and ML under limited supervision. With the extensive advancement in data collection and information availability, we focus on leveraging available learning resources to maximize learning potential and compensate for arising limitations. In addition, given the data-driven nature of ML, we aim to address key data limitations that significantly impact the quality of ML systems. Moreover, the wide adoption of ML applications raises significant ethical concerns, which we address as part of our research in artificial intelligence (AI) ethics policy and education. To address common data limitations, we develop robust models insensitive to noise and insufficient training supervision. Furthermore, to leverage available learning resources, we design multisource learning mechanisms in three broad categories:
[1] Graph-based Learning to consider data dependencies.
[2] Knowledge-Enhanced Learning to incorporate knowledge about the domain, known patterns or structures in the data, data available in various modalities, or multiple information channels in the form of data or ensembles of models.
[3] Data-driven System Modeling to integrate a “learning-from-data” ML approach into conventional “system-model-based” control methods to solve problems for complex systems for which classic control methods are not adaptable or scalable.

Topics

• Natural Language Information Extraction
• Time-Series Prediction
• Control

Applications

• Computer Vision
• Graph Learning
• AI Ethics
• Cyber Security
• Healthcare
• Transportation

Sponsors

• IARPA (Intelligence Advanced Research Projects Activity)
• NSF (National Science Foundation)
• DOE (Department of Energy)
• NIH iTHRIV (National Institute of Health integrated Translational Health Research Institute of Virginia)
• CCI (Commonwealth Cyber Initiative)
• Siemens
• Adobe
• eBay
• Virginia Tech
• Proctor & Gamble
• Xerox
• PARC (Palo Alto Research Center)
• DARPA (Defense Advanced Research Projects Agency)

Selected top venues where our research is published

Top Machine Learning Journals and Conferences

• Journal of Machine Learning Research (JMLR)
• ACM Transactions on Knowledge Discovery from Data (TKDD)
• ACM Transactions on Intelligent Systems and Technology (TIST)
• IEEE Transactions on Big Data
• IEEE Transactions on Knowledge and Data Engineering (TKDE)
• ACM Conference on Artificial Intelligence (AAAI)
• ACM SIGKDD Conference on Knowledge Discovery & Data Mining (KDD)
• ACM International World Wide Web Conference (WWW)
• ACM International Conference on Information and Knowledge Management (CIKM)
• ACM International Conference on Neural Information Processing (ICONIP)
• ACM Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD)
• IEEE International Conference on Big Data
• IEEE International Joint Conference on Neural Networks (IJCNN)
• Conference on Empirical Methods in Natural Language Processing (EMNLP)
• ACL International Conference on Computational Linguistics, Language Resources and Evaluation (COLING-LREC)

Top Control Journals and Conferences

• Systems and Control Letters
• IEEE Control Systems Letters
• IEEE Automatic Control Conference (ACC)
• IEEE European Control Conference (ECC)
• IEEE Conference on Decision and Control (CDC)

Other Top Journals and Conferences

• Transportation Research Record (TRR)
• IEEE Transactions on Technology and Society
• IEEE Frontiers in Education Conference (FIE)
• ACM International Conference on Pervasive and Ubiquitous Computing (UbiComp)
• ACM International Conference on Mobile and Ubiquitous Multimedia (MUM)
• ACM Pervasive Health Conference (PHC)
• IEEE Conference on Magnetism and Magnetic Materials (MMM)
• IEEE EMBS Conference on Neural Engineering (NER)