Skip to Main Content

IS Department Research Seminar Series

 

 

 

Please join us as we bring you a series of talks highlighting department research and topical issues in our field.

Scheduled Seminars

Title: Deep Learning Models for Healthcare and Oncology
Time: Thursday, October 17, 12-1 pm
Location: ITE 459
Speaker: Sanjay Purushotham
RSVP on Facebook

Abstract:
It is widely believed that machine learning and artificial intelligence techniques will substantially change healthcare industries. Even though recent developments in machine learning, particularly deep learning, has achieved success in many applications, such as computer vision, natural language processing, speech recognition, and so on, healthcare applications pose many significantly different challenges to existing deep learning models. Examples include but are not limited to interpretations for prediction, heterogeneity in data, missing values, multi-rate multi-resolution data, big and small data, and privacy issues. In this talk, I will discuss a series of problems in healthcare that can benefit from deep learning models, the challenges as well as recent advances in addressing those. I will also present the findings from our recent foray of applying deep learning models to oncology research.

Bio:
Sanjay Purushotham is an Assistant Professor in the Department of Information Systems at the University of Maryland, Baltimore County (UMBC). Before joining UMBC, he was a Postdoctoral Scholar Research Associate in the Department of Computer Science at the University of Southern California (USC). He obtained his M.S and Ph.D. in Electrical Engineering from USC. His research interests are in machine learning, data mining, optimization theory, statistics, computer vision, and its applications to healthcare & bioinformatics, oncology, and multimedia data analytics. Recently, he has been developing deep learning frameworks to model healthcare data and to predict survival outcomes for cancer patients. He has produced more than 30 publications and has won the best paper and best poster awards at international conferences.

 

Title: Big Data Challenges and Opportunities in Climate Science
Time: Thursday, October 24, 12-1 pm
Location: ITE 459
Speaker: Jianwu Wang
RSVP on Facebook

Abstract:
With the developments of remote sensing techniques and global climate models, we are facing astronomically growing amount of data. For instance, MODIS data from satellites Terra and Aqua has generated around 500 TB raw data and petabyte processed data. Simple data processing tasks on top of these large volume dataset, such as average and standard deviation calculation, have become difficult on single machines. Meanwhile, data-driven analytics techniques, including deep learning, have provided new opportunities to study climate as a complement or replacement of physics-based simulation approaches. The talk will provide an overview of the big data challenges and opportunities in climate science and related research done at the Big Data Analytics Lab at UMBC.

Bio:
Dr. Jianwu Wang is an Assistant Professor at the Department of Information Systems, University of Maryland, Baltimore County (UMBC). He leads the Big Data Analytics Lab at UMBC and is also an affiliated faculty at the Joint Center for Earth Systems Technology (JCET), UMBC. His research interests include Big Data Analytics, Scientific Workflow, Distributed Computing, Service Oriented Computing. He has published 80+ papers with more than 1200 citations. He is/was associate editor or editorial board member of four international journals, co-chair of four related workshops. He is also program committee member for over 30 conferences/workshops, and reviewer of over 15 journals or books. Since joining UMBC in 2015, he has received multiple grants as PI funded by NSF, NASA, DOE, State of Maryland, and Industry. His current research interests include Big Data Analytics, Distributed Computing and Scientific Workflow with application focuses on climate and manufacturing.

 

 

Past Seminars

Title: Bayesian Modeling of Intersectional Fairness: The Variance of Bias
Time: Thursday, September 19, 12-1 pm
Location: ITE 459
Speaker: Jimmy Foulds
RSVP on Facebook

Abstract:
With the rising influence of machine learning algorithms on many important aspects of our daily lives, there are growing concerns that biases inherent in data can lead the behavior of these algorithms to discriminate against certain populations. Informed by the framework of intersectionality from the Humanities literature, we propose mathematical definitions of AI fairness that aim to ensure protection along overlapping dimensions including gender, race, sexual orientation, class, and disability. We prove that our fairness criteria behave sensibly for any subset of the set of protected attributes, and we illustrate links to differential privacy. Finally, we present a Bayesian probabilistic modeling approach for the reliable, data-efficient estimation of fairness with multi-dimensional protected attributes. Experimental results on criminal justice, census, and synthetic data demonstrate the utility of our methodology, and show that Bayesian methods are valuable for the modeling and measurement of fairness in an intersectional context.

Bio:
Dr. James Foulds (Jimmy) is an Assistant Professor in the Department of Information Systems at UMBC. His research interests are in both applied and foundational machine learning, focusing on probabilistic latent variable models and the inference algorithms to learn them from data. His work aims to promote the practice of probabilistic modeling for computational social science, and to improve AI’s role in society regarding privacy and fairness. He earned his Ph.D. in computer science at the University of California, Irvine, and was a postdoctoral scholar at the University of California, Santa Cruz, followed by the University of California, San Diego. His master’s and bachelor’s degrees were earned with first class honours at the University of Waikato, New Zealand, where he also contributed to the Weka data mining system.

 

 

Title: Center for Accelerated Real Time Analytics (CARTA): Next Generation Research
Time: Thursday, September 26, 10-11 am
Location: ITE 459
Speaker: Karuna Joshi
Speaker’s website: http://karuna.informationsystems.umbc.edu
RSVP on Facebook

Abstract:
CARTA is an NSF sponsored Industry-University Cooperative Research Center (IUCRC) in UMBC focused on cutting edge inter-disciplinary research in real time analytics using next generation accelerated hardware. This center, lead by UMBC, also includes Rutgers University, New Brunswick, North Carolina State University (NCSU) and Rutgers University, Newark. Our international partner is Tel Aviv University. The overall Center Director is Dr. Yelena Yesha and the UMBC Site Director is Dr. Karuna P Joshi. This interactive talk will cover information about CARTA’s resources and research projects.

Bio:
Karuna Pande Joshi is an Assistant Professor of Information Systems at UMBC and UMBC Site Director of CARTA. She also directs the Knowledge Analytics Cognitive and Cloud (KnACC) Lab. Her research focus is in the areas of Data Science, Cloud Computing, Data Security and Privacy and Healthcare IT systems. She has published over 50 papers and her research is supported by ONR, NSF, DoD, GE Research and Cisco. She teaches courses in Big Data, Database Systems Design and Software Engineering. She received her MS and PhD in Computer Science from UMBC, where she was twice awarded the IBM PhD Fellowship, and her Bachelors in Computer Engineering from the University of Mumbai, India. Dr. Joshi also has extensive experience of working in the industry primarily as an IT Program/Project Manager at the International Monetary Fund.

 

 

Title: Information Extraction: From General Domain to Specific Domain
Time: Thursday, October 10, 12-1 pm
Location: ITE 459
Speaker: Arpita Roy
RSVP on Facebook

Abstract:
Information extraction (IE) is the task of automatically extracting structured information from unstructured text. IE is an important NLP task with many practical applications including business intelligence, scientific research, healthcare records management, financial investigation and social media analysis. Due to the difficulty and diversity of the problem, we divide IE into two different categories; general domain IE and domain specific IE. As each of these domains has its’ own unique characteristics, we propose two different noble IE approaches for each domain. For general domain we focus on leveraging knowledge from existing IE systems and build a superior IE model. For specific domain like cyber security we propose using structured domain knowledge for better IE. In both domain we use machine learning based IE techniques.

Bio:
Arpita Roy is a PhD student at the Department of Information Systems at University of Maryland, Baltimore County, working with Dr. Shimei Pan in the Text Mining and Social Media Analytics Lab. Her research interests include Natural Language Processing, Machine Learning and Text Mining. Her current research focus is developing machine based Information Extractions models for different domains.