The application deadline is approaching!

Following up on our earlier message announcing the summer 2024 NextGenPop program at Duke University with an application deadline on February 5, 2024, we are now writing to announce a VIRTUAL INFORMATION SESSION to be held from 7-8pm on January 23, 2024 at this zoom link. Speakers will include Giovanna Merli and Jay Pearson at Duke, Emily Agree at Johns Hopkins and Marcy Carlson at the University of Wisconsin-Madison.  These speakers will briefly introduce the program and answer questions, along with a few former NextGenPop trainees who will share their experiences with your students. 

NextGenPop is a summer training initiative funded by the National Institute of Child Health and Development (NICHD) designed to identify, recruit and mentor undergraduate students who aspire to careers in the population research and population health fields. The program offers an introduction to population health sciences, through the lens of pressing contemporary issues, including race and income inequalities, health inequality, immigration, and family change. The program features a two-week, in-person, on-campus summer experience and subsequent virtual components focused on research and professional development, as well as opportunities for mentorship and networking at the Annual Meeting of the Population Association of America (PAA). The program provides a stipend, room, board, and travel support. In Summer 2024, the program will be hosted by the Duke University Population Research Institute at Duke University from June 2 to June 15, 2024.

A primary focus of NextGenPop is to increase representation from socially minoritized groups. Honoring this goal entails identifying, recruiting, and mentoring the most promising students currently underrepresented in population science. We have committed to creating a welcoming environment promoting a sense of belonging for all participants. Program leadership has prioritized: 1. selecting instructors and mentors from a range of broadly representative groups;  2. establishing, developing, and maintaining authentic relationships with faculty and administrators who have studied, trained and served at undergraduate minority-serving institutions; 3. honoring the belief that those whose lives reflect relevant phenomena of interest should have a seat at the research, instructional and policy making tables; and 4. practicing cultural humility as we continue learning from each other through a shared commitment to offset historical oppression and contemporary marginalization that restrict and constrain participation by members of socially minoritized populations. Finally, we place absolute value on dignity and respect for all participants whether trainees, staff, mentors, instructors, or administrators as we collectively learn how best to address difficult and challenging social issues disproportionately impacting the well-being and health of select population groups. We look forward to receiving your application to our program. 

The DEADLINE for application to the program is February 5, 2024. You and your students are invited to join us for the information session by clicking on this zoom link https://duke.zoom.us/j/96327878820.

For 2024, M.O.S.T. is seeking proposals for two grant programs. Learn more below. Note that all institutional and faculty grants awarded to UMBC must be routed to the Office of Sponsored Programs prior to submission. Contact your unit's business manager or proposal development staff to discuss your application.

CAHSS faculty can request assistance from CARAT (Creative Achievement and Research Administration Team). 

Faculty OER Grant Program
The Maryland Open Source Textbook (M.O.S.T.) initiative is seeking proposals for its Faculty OER Grant Program, which supports faculty efforts to increase access, affordability, and achievement for students at Maryland’s postsecondary institutions through the incorporation of open educational resources (OER) into teaching practice.

Available grants in 2024:

• Adopt/Adapt OER ($1,000 grant + Participation in the Lumen Circles Community of Practice)
• Create/Revise and Publish OER ($2,000 grant + Participation in the Rebus Textbook Success Program)
• OLI STEM Courseware Grant - Use & Evaluate Track ($1,500 grant + Participation in the NSF-Funded Carnegie Mellon Open Learning Initiative (OLI) STEM Courseware Project) NOTE: Available for 2-year colleges only
• OLI STEM Courseware Grant - Use & Improve Track ($5,000 grant + Participation in the NSF-Funded Carnegie Mellon Open Learning Initiative (OLI) STEM Courseware Project) NOTE: Available for 2-year colleges only

Institutional OER Grant Program
The Maryland Open Source Textbook (M.O.S.T.) initiative is seeking proposals for the Institutional OER Grant Program, which supports strategic institutional efforts to increase access, affordability, and achievement for students at Maryland’s postsecondary institutions through more systemic incorporation of open educational resources (OER) into teaching practice.

For 2024, selected institutions will be awarded $20,000 grants to support significant advances in OER adoption, scaling, and sustainability in support of student success. In addition, grantees will receive strategic planning and implementation support for their projects from M.O.S.T. in the form of consultations and regular monthly check-ins.
 

All proposals must be submitted by 11:59 p.m. ET on Friday, March 1, 2024
 

Informational Webinar
Join us on Monday, January 22, 2024 from 2:00 p.m. – 3:00 p.m. ET for an Informational Webinar to learn more about the grant requirements and expectations for successful grants. We will also answer any questions about the grants at that time. Register for the Informational Webinar
 

Please feel free to share this information with interested faculty and staff. If you have any questions, please reach out to us at most@usmd.edu 

Thank you for your support in increasing access, affordability, and achievement for all Maryland higher education students through open educational resources (OER).
 

Towards Democratizing AI: Scaling  and Learning (Fair) Graph Representations in an Implementation Agnostic Fashion

Srinivasan Parthasarathy, Ohio State University

Abstract: Recently there has been a surge of interest in designing graph embedding methods. Few, if any, can scale to a large-sized graph with millions of nodes due to both computational complexity and memory requirements. In this talk, I will present an approach to redress this limitation by introducing the MultI-Level Embedding (MILE) framework – a generic methodology allowing con-temporary graph embedding methods to scale to large graphs. MILE repeatedly coarsens the graph into smaller ones using a hybrid matching technique to maintain the backbone structure of the graph. It then applies existing embedding methods on the coarsest graph and refines the embeddings to the original graph through a graph convolution neural network that it learns. Time permitting, I will then describe one of several natural extensions to MILE  -  in a distributed setting (DistMILE) to further improve the scalability of graph embedding or mechanisms -  to learn fair graph representations (FairMILE).  

The proposed MILE framework and variants (DistMILE, FairMILE), are agnostic to the underlying graph embedding techniques and can be applied to many existing graph embedding methods without modifying them and is agnostic to their implementation language. Experimental results on five large-scale datasets demonstrate that MILE significantly boosts the speed (order of magnitude) of graph embedding while generating embeddings of better quality, for the task of node classification. MILE can comfortably scale to a graph with 9 million nodes and 40 million edges, on which existing methods run out of memory or take too long to compute on a modern workstation. Our experiments demonstrate that DistMILE learns representations of similar quality with respect to other baselines while reducing the time of learning embeddings even further (up to 40 x speedup over MILE). FairMILE similarly learns fair representations of the data while reducing the time of learning embeddings.