| NSF Org: |
CNS Division Of Computer and Network Systems |
| Recipient: |
|
| Initial Amendment Date: | September 12, 2018 |
| Latest Amendment Date: | September 12, 2018 |
| Award Number: | 1831980 |
| Award Instrument: | Standard Grant |
| Program Manager: |
Marilyn McClure
mmcclure@nsf.gov (703)292-5197 CNS Division Of Computer and Network Systems CSE Directorate for Computer and Information Science and Engineering |
| Start Date: | October 1, 2018 |
| End Date: | September 30, 2023 (Estimated) |
| Total Intended Award Amount: | $466,515.00 |
| Total Awarded Amount to Date: | $466,515.00 |
| Funds Obligated to Date: |
|
| History of Investigator: |
|
| Recipient Sponsored Research Office: |
3100 CLEBURNE ST HOUSTON TX US 77004-4501 (713)313-7457 |
| Sponsor Congressional District: |
|
| Primary Place of Performance: |
3100 Cleburne Houston TX US 77004-4501 |
| Primary Place of
Performance Congressional District: |
|
| Unique Entity Identifier (UEI): |
|
| Parent UEI: |
|
| NSF Program(s): | HBCU-EiR - HBCU-Excellence in |
| Primary Program Source: |
|
| Program Reference Code(s): |
|
| Program Element Code(s): |
|
| Award Agency Code: | 4900 |
| Fund Agency Code: | 4900 |
| Assistance Listing Number(s): | 47.070 |
ABSTRACT
This project will be implemented by an interdisciplinary research team from two neighboring historical Black Colleges and Universities (HBCUs), Prairie View A&M University (PVAMU) and Texas Southern University (TSU), with the goals of strengthening the theoretical foundation of big data analytics and developing a novel deep learning software package based on this enhanced foundation. Challenges around big data impacts many areas and encourages exciting further investigation to understand the complex requirements of real-world applications.
Specifically, this team aims to 1) improve the understanding and explainability of deep neural networks with the quantum theory of modern physics; 2) enhance the mathematical foundation of deep neural networks; 3) increase the computation efficiency of the deep learning training process with new algorithms that will scale; and 4) implement the deep learning research innovations into a new deep learning software package to deploy in cloud computing and High-Performance Computing platforms.
This project will aid the research in a wide range of areas of applications, from academic studies to the oil and gas industry and the military by offering deep learning models and improved computational efficiency and scalability. Students will benefit from the opportunity to contribute to the forefront of science research and technology. The project will broaden participation by opening stimulating research opportunities to a diverse group of underrepresented minority students. The two campuses and the members of the interdisciplinary team complement each other; by exchanging and sharing expertise and students the collaboration therefore has a beneficial and synergistic effect on multiple levels.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
PUBLICATIONS PRODUCED AS A RESULT OF THIS RESEARCH
Note:
When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external
site maintained by the publisher. Some full text articles may not yet be available without a
charge during the embargo (administrative interval).
Some links on this page may take you to non-federal websites. Their policies may differ from
this site.
PROJECT OUTCOMES REPORT
Disclaimer
This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.
Researchers from Prairie View A&M University and Texas Southern University, two major Historical Black Colleges and Universities in Houston area joined forces to make advances in theoretical and mathematical foundations of deep learning. Joint activities of this project, including workshops and intensive undergraduate summer research bootcamps, have stirred interest of underrepresented minority student in STEM with an enrichment of the learning environment in Data Science at both universities. This program has started building capacity in big data analytics and advanced deep learning at these campuses by hiring postdoctoral fellows and having regular meetings that stimulated active exchange of ideas. A platform of collaboration and teaching based on a jupyterhub server was build and adopted for training students in research, as well as for use in computational class work. The collaboration continues through other projects.
More specifically, the outcomes from this project include the following.
Research. This project obtained notable results towards the main goal of understanding and increasing explainability and efficiency of deep neural networks. Researchers designed methods, implemented algorithms, and constructed software for infusing the existing agnostic deep learning techniques with constrains extracted from the physics of the concrete applications considered, such as the classification of seismic activity as natural or anthropogenic, or detection of neurological diseases based on electroencephalogram (EEG) signals. As an outcome of this effort three new collaborative proposals were submitted to further explore the area of scientific machine learning and deepen the connections among researchers from both institutions.
Results obtained in this project demonstrated ways of improving the efficiency of machine learning by investigating mathematical foundation of backpropagation auto-differentiation methods, and by systematic exploration of structures and topologies of deep neural networks.
A total of 12 papers published were supported or partially supported by the grant, and 10 conference presentations were delivered.
Student trained. A total of 23 undergraduate and high school students have participated in four 10-week summer research program SURE-DS (Summer Undergraduate Research in Data Science) in 2019, 2020 (virtual), 2021 (virtual) and 2023. Students received regular lectures in python programming and machine learning, were trained in literature search and conducted research projects. At the end of each summer, the students produced posters and/or manuscripts and gave oral presentation.
Annual joint workshops and student organization. Three annual workshops were successfully hosted and have attracted more and more researchers every year. A student’s organization for data science (Tiger’s Data Science Club), led by several summer trainees, has born out of lunch network discussions in 2023. The club has successfully organized two events with featured speakers from Oracle and Google. The team has facilitated to build partnership among the student organization between the two sites.
Developed Training material for Data science. A set of lectures, jupyter notebooks and projects were developed that can be reused for courses in data science and future training activities. The materials include Introduction to Python, Introduction to Data Science and Machine Learning, and also include slides for workshops on how to read scientific research papers, how to conduct research, how to prepare poster and present research finding and how to write up manuscripts.
Jupyterhub server. A platform of jupyterhub sever was developed to facilitate training, teaching and research. For various reasons, students in general have trouble in properly installing programming languages such as Python, R in their own computers. This server provides an easy solution for addressing this issue. More importantly, with this platform, non-tech savvy faculties, such as most math professors, now have the freedom to embed computational components in many math courses, such as Linear Algebra, Calculus, Differential Equations, Numerical Analysis, Statistics, and Machine Learning. Up to now, the platform has been regularly used in several math and engineering courses.
Curriculum development. The PIs have taken the lead to develop courses and programs in data science on campus. New courses including Introduction to Data Science and Machine Learning will be offered first time in the Spring of 2024. A minor in data science has started to be implemented in the Math department since the spring of 2023.
New Grant proposals submitted. Due to this project the team has enhanced the knowledge, skills and research capability in deep neural network and data science in general and have developed new projects in scientific machine learning and have started collaborations with researchers from various backgrounds. The team were invited to join many grant proposals and have led five new grant proposals to different funding agencies.
Last Modified: 12/03/2023
Modified by: Yunjiao Wang
Please report errors in award information by writing to: awardsearch@nsf.gov.
