UMBC receives NSF support to enhance data science courses, research and student experiences

Two NSF awards will enhance data science courses, research, and student experiences

Data science has rapidly grown at UMBC, and faculty are now working to enhance data science courses, research, and student experiences. The goal is to ensure they are inclusive, empowering, and effective in preparing students to tackle the urgent problems our society is working to solve, and can scale up to meet student and workforce demands.

Faculty in both information systems, and computer science and electrical engineering have recently received two grants from the National Science Foundation to conduct research toward this aim.

Making data science more inclusive

NSF awarded funding to a team of researchers at UMBC; the University of California, Berkeley; and Mills College in California through the Improving Undergraduate STEM Education program. This funding will support the Undergraduate Data Science at Scale project at UMBC, including the development and implementation of a unique data science education program for undergraduate students in STEM and non-STEM disciplines, says Vandana Janeja, associate professor and chair of information systems (IS).

The data science field is relatively new as compared to the more established computing education, Janeja explains, and there are very few studies examining how these topics are taught to students. “This project will generate new knowledge about a data science curriculum and pedagogy designed to promote learning among diverse undergraduate students, many from groups underrepresented in STEM,” she says.

This novel approach to teaching data science will also “empower students as generators of new knowledge rather than passive recipients of existing information,” Janeja explains.

Another component of the project is the data scholars program, which will include students from traditionally underrepresented groups in STEM fields.

With these changes, Janeja anticipates that data science at UMBC will continue to expand. She’s working with UMBC’s Division of Information Technology to explore how UMBC, and other universities, will need to adapt and scale up offerings to meet the changing needs of students over time. “The findings will drive a community transformation in undergraduate data science education that can scale with student demand, and ultimately broaden participation in data science across multiple and diverse institutional settings.”

Janeja and UMBC colleagues are excited to develop a model that can have a nationwide impact, bringing new students into the field, and shaping how they approach work in data science. She expects that this work will set a foundation for colleges across the country looking to implement data science programs and better support the learning of data science students.

High-impact, team-based student research

Outside of the classroom, undergraduate students in computer science, information systems, and business technology administration will have the opportunity to work with government agencies in Baltimore City to tackle real challenges through a new NSF-funded program. Aryya Gangopadhyay, professor of information systems, has received support for the new program through the Data Science Corps under NSF’s Harnessing the Data Revolution (HDR) initiative.

HDR is one of NSF’s 10 “Big Ideas”: bold, visionary, national-scale activities to open up new frontiers in science and engineering. This program allows researchers to answer fundamental questions through new modes of data-driven discovery, Gangopadhyay explains. On this project, he will work with colleagues and students to collect and analyze data for projects that seek to improve Baltimore residents’ quality of life.

Gangopadhyay will partner with UMBC faculty including Anupam Joshi, professor and chair of computer science and electrical engineering (CSEE); Tim Oates, professor of CSEE; Nirmalya Roy, associate professor of IS; and Sanjay Purushotham, assistant professor of IS. The UMBC team will collaborate with faculty at Bowie State University, Towson University, and the University of Baltimore. Gangopadhyay and his team will work with UMBC’s Faculty Development Center to evaluate student learning outcomes for this project.

“The goal of the project is to develop a team-based data science program for undergraduate students in computing,” explains Gangopadhyay. Both undergraduate and graduate students will contribute to this research, gaining hands-on experience with the complexity of addressing urban infrastructure challenges, such as traffic congestion.

Students will also examine a range of ethical considerations, including data privacy, as they process information. Street sensors, for example, can collect sensitive information on peoples’ patterns of daily life. Students will come to better understand their role as researchers in protecting privacy, and other ethical considerations, as they cull through the data, says Gangopadhyay.

“Data science is poised to change the world by improving the quality of life through smart technologies,” explains Gangopadhyay. “Our students will play a part in bringing about some of these changes. Through their projects, students will develop analytical and coding skills and learn how to collaboratively work in real life projects with industry, government and academia, under the guidance of faculty mentors.”

Adapted from a UMBC News article by Megan Hanks. Banner image: A student using a computer. Photo by Marlayna Demond ’11 for UMBC.

talk: Ian Blumenfeld on Interactive Proof Assistants for Verification, Fri 1/31

The UMBC Cyber Defense Lab presents

Interactive Proof Assistants for Verification

Ian Blumenfeld
Principal Research Mathematician
Two Six Labs   

12:00-1:00 pm Friday,  31 January 2020, ITE 227

Many advances have been made in software and hardware assurance using automated tooling.  Constraint-based solving tools like SAT and SMT solvers have proved very useful proving functional correctness in the world of software, while the hardware world relies heavily on the use of industrial-strength model checkers to provide formal verification of important properties like liveness and non-interference.  Sometimes, however, push-button tools are simply not enough. In this talk, we will discuss formal mathematical reasoning using interactive proof assistants, particularly Isabelle. While Isabelle is often thought of as a tool for checking the work of mathematicians, it is, in fact, a powerful engine for reasoning about software and hardware security.  We will work through an example of the verification of a multi-precision arithmetic software library using Isabelle. This talk is aimed at total beginners in the realm of automated theorem proving, and seeks to provide an overview of the fundamental techniques and ideas. 

Ian Blumenfeld is a Principal Research Mathematician at Two Six Labs.  He currently is the principal investigator of TwoSix’s efforts on the DARPA SafeDocs program, attempting to help do type-theoretic reasoning about document specification formats.  He is a former employee of Apple where he worked on the formal verification team, ensuring the security of the iPhone SEP chip. He has done extensive work verifying cyber-physical systems at Johns Hopkins APL.  Mr. Blumenfeld’s interest in formal methods began with his time working as an Applied Research Mathematician in NSA’s Research Directorate. He’s also a pretty good swing dancer.

Host: Alan T. Sherman, 

Support for this event was provided in part by the National Science Foundation under SFS grant DGE-1753681. 

The UMBC Cyber Defense Lab meets biweekly Fridays.  All meetings are open to the public.

JHU/APL CIRCUIT internship program information session, 3pm Fri 1/31

JHU/APL CIRCUIT internship program information session

3:00-4:00 pm Friday, 31 January 2020

ITE 459, UMBC

There will be a special information session on the JHU/APL CIRCUIT internship program from 3:00 pm to 4:30 pm on Friday, 31 January 2020 in room ITE 459.

This session is for undergraduates who want to spend their summer (June through August) getting paid to do mentored research at the Johns Hopkins University Applied Physics Lab. The research areas include AI, data science, cybersecurity, precision medicine, and planetary exploration.

Interns selected for the program will do mission-oriented research on-site at JHU/APL in Laurel MD mentored by STEM professionals. There will also be year-round opportunities for engagement and enrichment. The selection for an internship will be based on a combination of potential, need and commitment.

Email or with questions.

Prof. Tim Oates Receives IARPA Contract under TrojAI Program

Prof. Tim Oates Receives IARPA funding to protect AI models from malicious actors

Professor Tim Oates, working with ARM Research, just received a contract from the Intelligence Advanced Research Projects Activity (IARPA) under the TrojAI program.  The goal of the work is to detect trojans hiding in deep neural networks (DNNs). 

DNNs are widely used in industry to solve problems like recognizing stop signs in self-driving cars and finding low and slow attacks on corporate intranets.  But there are ways of attacking DNNs during training to implant a trojan, an input pattern that can cause the network to do the wrong thing.  For example, one could poison the stop sign recognizer so that when a yellow sticky note is placed on the stop sign the network sees it as a speed limit sign. 

Professor Oates will explore ways of examing trained DNNs, without access to the data on which they were trained, to determine if they are hiding trojans.  His work will help protect systems that use DNNs from malicious actors.

Science Unscripted: Conversations with AI Experts, 5-8:00pm 29&30 Oct 2019, UMBC

On October 29 and 30 the National Science and Technology Medals Foundation will host Science Unscripted: Conversations with AI Experts, two early evening events at UMBC from 5:00 to 8:00pm that bring together AI experts to discuss how AI will impact our lives. The events will be held in the Fine Arts Recital Hall with doors open at 5:00 prior to the 5:30 start and will conclude with a reception starting at 7:00pm with food and drinks. Both events are free, but registration is requested.

These events are a part of the NSTMF’s Science Unscripted program. Through the SU program, the Foundation is building an inclusive coalition of inspired STEM students. By highlighting voices often left unheard in the STEM community, we show audiences that there is no “right” way to be a trailblazer in science and technology. Each evening, attendees will have the chance to hear about the lives and experiences of the women and men dedicated to creating smart, socially conscious AI.

Tuesday, Oct. 29: Code-ifying AI is a a discussion about AI policy. A panel including UMBC Professor Cynthia Matuszek, Dr. José-Marie Griffiths and moderated by Rosario Robinson will examine what it will take to govern AI as well as the implications of incorporating AI into our everyday lives. Register on Eventbright.

Wednesday, Oct. 30: Decoding Bias in AI is a panel discussion about implicit bias and how we can create more socially conscious AI with UMBC Professor James Foulds, Loretta Cheeks, Emmanuel Johnson and moderator Deborah Kariuki. Implicit bias remains one of the most prevalent concerns about incorporating AI into the mainstream, and our panel is poised to deliberate the ethics and possible solutions to this issue. Register on Eventbright.

The events will be webcast live with closed-captions on Facebook, and the full event videos will be available on our YouTube channel afterward. Webcast audiences are encouraged to participate in the conversation using #ScienceUnscripted on Twitter, Facebook and Instagram.

Both events are no-cost, equal access (ADA compliant), and open to the public. Save your seat on Eventbrite for day one at Code-ifying AI and for day two at Decoding Bias in AI.

TALK: Computer Aided Assessment of Computed Tomography Screenings

UMBC ACM Chapter Talk

Computer Aided Assessment of Pulmonary Nodule Malignancy in from Low Dose Computed Tomography Screenings

Professor David Chapman, CSEE, UMBC

11:30–12:30, Friday 11 October 2019, ITE 346, UMBC

We propose to develop a novel quantitative algorithm to estimate the probability of malignancy of pulmonary nodules from a time series of successive LDCT screenings in patients with a high risk of developing lung cancer. Lung cancer kills approximately 200,000 Americans annually and is responsible for 25% of all cancer-related deaths. Imaging with Low Dose Computed Tomography (LDCT) has been proven to reduce Non-Small Cell Lung Cancer (NSCLC) mortality by 20% and has become standard guidelines (NLST 2011a,b). These new clinical guidelines have led to hospitals, including Mercy Medical Center in Baltimore, to collect an abundance of LDCT images of high risk individuals since 2014. These LDCT images along with additional CT/biopsy and PET/CT images collected by Mercy hospital in Baltimore have now been organized into an IRB exempt clinical research dataset to use anonymous radiology imagery for the purpose of training and evaluation of improved Computer Aided Diagnosis (CAD) algorithms. Imaging biomarkers including cross-sectional diameter, calcification patterns, irregular margins, wall thickness all of which are known to have discriminating power to differentiate benign and malignant pulmonary nodules. Furthermore, temporal changes in the size and biomarker characteristics of pulmonary nodules over multiple images are also highly informative and yield greater ability to differentiate malignancy. The proposed CAD algorithm will be capable of detecting and quantifying temporal changes of imaging biomakers in order to estimate malignancy probability. The algorithm will make use of convolutional neural networks for feature extraction as well as recurrent neural networks to analyze the temporal changes in extracted features. The Mercy hospital dataset contains approximately 30,000 chest CT images. Training of the algorithm will incorporate semi-supervised learning using chest CT images from Mercy as well as the public portion of the NLST dataset. A fraction of the Mercy images will be designated for evaluation of the sensitivity and specificity of the proposed algorithm for determining nodule malignancy. Pulmonary nodules remain a challenging area for clinical management decision-making, and improved analysis of malignancy including temporal changes of imaging biomarkers have the potential to reduce patient morbidity and mortality through earlier and more accurate diagnosis.

talk: Three Related Takes on Investigating Human-Like Intelligence

Lockheed Martin Distinguished Speaker Series

Three Related Takes on Investigating Human-Like Intelligence:
Cognitive Architectures, a Common Model of Cognition, and Dichotomic Maps

Dr. Paul S. Rosenbloom

Professor of Computer Science and Director of Cognitive Architecture Research, Institute for Creative Technologies
University of Southern California

1:00-2:00pm Friday, 11 October 2019, ITE 325b, UMBC

This talk explores a trio of related takes on how to investigate the nature of human-like intelligence. The first concerns cognitive architectures – implemented models of the fixed structure and processes that yield natural and artificial minds – with a drill down to Sigma, an attempt at a deep synthesis across what has been learned over the past four decades on (what started as) high-level symbolic cognitive architectures versus the low-level graphical/network technologies of probabilistic graphical models (such as Bayesian networks) and neural networks. The second concerns a more abstract attempt at specifying a Common Model of Cognition that yields an evolving community consensus over what must be part of any cognitive architecture for human-like intelligence. The final take concerns an even more abstract (and speculative) attempt at understanding more deeply the space of approaches to intelligence – framed as maps resulting from cross products among core cognitive dichotomies – along with how such maps may help to understand and structure the capabilities required for (human-like) intelligence.

Paul Rosenbloom is a professor of computer science in the Viterbi School of Engineering at the University of Southern California (USC) and director for cognitive architecture research at USC’s Institute for Creative Technologies (ICT). He was a member of USC’s Information Sciences Institute for two decades, ending as its deputy director, and earlier was on the faculty at Carnegie Mellon University and Stanford University (where he had a joint appointment in Computer Science and Psychology). His research concentrates on cognitive architectures (models of the fixed structures and processes that together yield a mind), the Common Model of Cognition (an attempt at developing a community consensus concerning what must be part of a human-like mind), and on computing as a scientific domain (understanding the computing sciences as akin to the physical, life and social sciences). He is a fellow of the Association for the Advancement of Artificial Intelligence (AAAI), the Association for the Advancement of Science (AAAS), and the Cognitive Science Society; and with J. Laird was recently awarded the Herbert A. Simon Prize for Advances in Cognitive Systems. He has served as councilor and conference chair for AAAI; chair of the Association for Computing Machinery Special Interest Group on Artificial Intelligence; and president of the faculty at USC.

New NSF award will help robots learn to understand humans in complex environments

Prof. Ferraro in UMBC’s Pi2 visualization laboratory talking to a virtual robot, modeled using a combination of Unity, ROS, and Gazebo. Image from a recent paper on this research..

New NSF project will help robots learn to understand humans in complex environments

UMBC Assistant Professor Cynthia Matuszek is the PI on a new NSF research award, EAGER: Learning Language in Simulation for Real Robot Interaction, with CO-PIs Don Engel and Frank Ferraro. Research funded by this award will be focused on developing better human-robot interactions using machine learning to enable robots to learn the meaning of human commands and questions informed by their physical context.

While robots are rapidly becoming more capable and ubiquitous, their utility is still severely limited by the inability of regular users to customize their behaviors. This EArly Grant for Exploratory Research (EAGER) will explore how examples of language, gaze, and other communications can be collected from a virtual interaction with a robot in order to learn how robots can interact better with end users. Current robots’ difficulty of use and inflexibility are major factors preventing them from being more broadly available to populations that might benefit, such as aging-in-place seniors. One promising solution is to let users control and teach robots with natural language, an intuitive and comfortable mechanism. This has led to active research in the area of grounded language acquisition: learning language that refers to and is informed by the physical world. Given the complexity of robotic systems, there is growing interest in approaches that take advantage of the latest in virtual reality technology, which can lower the barrier of entry to this research.

This EAGER project develops infrastructure that will lay the necessary groundwork for applying simulation-to-reality approaches to natural language interactions with robots. This project aims to bootstrap robots’ learning to understand language, using a combination of data collected in a high-fidelity virtual reality environment with simulated robots and real-world testing on physical robots. A person will interact with simulated robots in virtual reality, and his or her actions and language will be recorded. By integrating with existing robotics technology, this project will model the connection between the language people use and the robot’s perceptions and actions. Natural language descriptions of what is happening in simulation will be obtained and used to train a joint model of language and simulated percepts as a way to learn grounded language. The effectiveness of the framework and algorithms will be measured on automatic prediction/generation tasks and transferability of learned models to a real, physical robot. This work will serve as a proof of concept for the value of combining robotics simulation with human interaction, as well as providing interested researchers with resources to bootstrap their own work.

Dr. Matuszek’s Interactive Robotics and Language lab is developing robots that everyday people can talk to, telling them to do tasks or about the world around them. Their approach to learning to understand language in the physical space that people and robots occupy is called grounded language acquisition and is a key to building robots that can perform tasks in noisy, real-world environments, instead of being pre-emptively programmed to handle a fixed set of predetermined tasks.

talk: Tensor Decomposition of ND data arrays, 2pm 6/13 ITE325

Tensor Decomposition of ND data arrays

Prof. David Brie, University of Lorraine

2:00pm Thursday, 13 June 2019, ITE 325B, UMBC

The goal of this talk is to give an introduction to tensor decompositions for the analysis of multidimensional data. First, we recall some basic notions and operations on tensors. Then two tensor decompositions are presented: the Tucker decomposition (TD) and the Candecomp/Parafac decomposition (CPD). A particular focus is placed on the identifiability conditions of the CPD. Finally, various applications in biology are presented.

David Brie received the Ph.D. degree in 1992 and the Habilitation à Diriger des Recherches degree in 2000, both from Université de Lorraine, France. He is currently full professor at the Department of Telecommunications and Networking of the Institut Universitaire de Technologie, Université de Lorraine, France. He is editor-in-chief of the French journal “Traitement du Signal” since 2013 and will be co-general chair of the next IEEE CAMSAP 2019. His current research interests include vector-sensor-array processing, spectroscopy and hyperspectral image processing, non-negative matrix factorization, multidimensional signal processing, and tensor decompositions.

talk: Learning to Ground Instructions to Plans, 2:30 Thr 3/21, ITE346

Learning to Ground Natural Language Instructions to Plans

Nakul Gopalan, Brown University

2:30-3:30pm Thursday, 21 March 2019, ITE 346, UMBC

In order to easily and efficiently collaborate with humans, robots must learn to complete tasks specified using natural language. Natural language provides an intuitive interface for a layperson to interact with a robot without the person needing to program a robot, which might require expertise. Natural language instructions can easily specify goal conditions or provide guidances and constraints required to complete a task. Given a natural language command, a robot needs to ground the instruction to a plan that can be executed in the environment. This grounding can be challenging to perform, especially when we expect robots to generalize to novel natural language descriptions and novel task specifications while providing as little prior information as possible. In this talk, I will present a model for grounding instructions to plans. Furthermore, I will present two strategies under this model for language grounding and compare their effectiveness. We will explore the use of approaches using deep learning, semantic parsing, predicate logic and linear temporal logic for task grounding and execution during the talk.

Nakul Gopalan is a graduate student in the H2R lab at Brown University. His interests are in the problems of language grounding for robotics, and abstractions within reinforcement learning and planning. He has an MSc. in Computer Science from Brown University (2015) and an MSc. in Information and Communication Engineering from T.U. Darmstadt (2013) in Germany. He completed a Bachelor of Engineering from R.V. College of Engineering in Bangalore, India (2008). His team recently won the Brown-Hyundai Visionary Challenge for their proposal to use Mixed Reality and Social Feedback for Human-Robot collaboration.

Host: Prof. Cynthia Matuszek (cmat at umbc.edu)

1 2 3 4 5 … 7

©2024 University of Maryland Baltimore County Computer Science and Electrical Engineering Department
 1000 Hilltop Circle, ITE 325, Baltimore, Maryland 21250 | 410-455-3500 | ">
College of Engineering and Information Technology | Contact Us | Equal Opportunity | Consumer Information | Accessibility