Research and Seminars

Seminars, Computational Lunch, and DiSH

The data science program regularly participates in Computational Science Seminars & Lunchtime Computing hosted by the The Center for Scientific Computing and Data Science Research (CSCDR), and the Computer and Information Science (CIS) seminars series through CIS 599.

CIS seminars are typically held on Fridays at 3:00 PM and are announced through UMassD Announce Digest. 

CSCDR events are announced through UMassD Announce Digest and data science student/faculty email lists. A complete schedule of past and futures talks can be found on the CSCDR seminar page.

We also collaborate with the library through their Digital Scholarship Hub (DiSH) program.  DiSH holds regular hands-on training sessions covering a range of data science topics. DiSH events are announced on the library's Instagram and Facebook sites.

Data Science Research

Data science faculty and students are engaged in a wide range of research topics. Many of our faculty and students' computational work is carried out with the support of The Center for Scientific Computing and Data Science Research. Due to the interdisciplinary nature of our program, ongoing research activities are more fully described through the following links:

• Computer & Information Science Research: 
  • Bioinformatics and biomedical computing (Firas Khatib)
  • Neural computing, multi-agent systems (Haiping Xu)
  • Citizen Science and gamification (Firas Khatib)
  • Computer and information security (Yuchou Chang, Gokhan Kul)
  • Data visualization and computer vision (Ming Shao)
  • E-commerce (Haiping Xu, Shelley Zhang)
  • Data privacy and data security in cloud computing (Haiping Xu)
  • Computational Statistics, Missing Data Analysis, Trajectory Pattern Recognition, Pattern Validation and Visualization (Hua Fang)
  • Real-time Statistical/Machine Learning and Analytics for Big Data, Digital Health/Virtual Care and IoT applications (Hua Fang)
  • Cloud-based blockchains for big data storage (Haiping Xu)
  • Recommendation Systems (Shelley Zhang)
  • Machine/statistical learning, pattern recognition (Ming Shao, Iren Valova)
  • Data mining and text mining using deep learning (Haiping Xu)
  • Neural computing, multi-agent systems (Ramprasad Balasubramanian, Iren Valova, Shelley Zhang)
  • Decision-support with uncertainty (Shelley Zhang)
• Mathematics and Statistics:
  • Computational and statistical learning (Donghui Yan)
  • Distributed information sharing, inference and learning (Donghui Yan)
  • Mobile and digital health (Donghui Yan)
  • Data Science Education (Gary Davis, Saeja Kim)
  • Spatial Point Processes (Gary Davis)
  • Scientific machine learning (Yanlai Chen, Zheng Chen, Alfa Heryudono, Sigal Gottlieb)
  • Physics informed neural networks and universal differential equations (Scott Field,  Alfa Heryudono)
  • Deep matched filtering (Scott Field)
• Student capstone and practicum projects
  • Undergraduate capstone
  • Graduate thesis and practicum
  • Projects and hackathons sponsored by the BigDataClub
  • Data Science Capstone Day talks and abstracts

Tukey Rapid Production Server

The Tukey server, named in honor of New Bedford's own John Tukey, is a high-end computational resource for data science faculty and students to carry out computationally demanding AI and Big Data problems. Tukey was purchased by the data science program with additional support from the CSCDR, College of Engineering, College of Arts and Sciences, Department of Computer & Information Science, and Department of Mathematics.

• Hardware specs: 64 AMD Epyc cores, 1 TB of DDR3 RAM, 20 TB of storage, and two NVIDIA A100 GPUs with 80 GB of RAM
• Software specs: Ubuntu 20.04, JupyterHub server
• How do I use Tukey? Tukey runs a JupyterHub server. The machine can be accessed through a JupyterHub URL, and from here you can use Jupyter notebooks, gain terminal access, and run code as you normally would. If you are unfamiliar with JupyterHub, there are many excellent YouTube videos introducing the basics for getting started.
• How do I access Tukey? Tukey is primarily meant for research computing, which can include student capstone, practicum, or thesis projects. Data Science faculty who would like access should contact the data science co-directors. Students who would like access should have a faculty sponsor -- this could be your capstone teacher, practicum/thesis advisor, or research supervisor. Faculty interested in using this machine in a course should first discuss their plans with the data science co-directors.
• Have you used Tukey in your work? Please acknowledge this machine as follows: "A portion of computational work of this project was performed on the Tukey server at UMassD, which is supported by the Data Science program and the Center for Scientific Computing and Data Science Research." 

Open Data LibGuide

.