South Big Data Hub Education & Workforce Working Group, 16 Dec 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Education & Workforce WG: NIH AIM-AHEAD Award Overview: Lead the Data Science Training Core

Description

November 2021
South Hub Education & Workforce Working Group
Presenter: Dr. Legand Burge; Howard University

A

All right yeah, uh thank you uh renata! So um yes, uh so um again, my name is uh lee burge, professor of computer science at howard university and I'm the current pi of the data science training core, which is part of a larger initiative that renata talked about called aim ahead, which stands for artificial intelligence, machine learning consortium to advance health, equity and research diversity. I won't spend too much time on this because she did give you a nice overview, but just conceptually so you can picture, have a you know. Conceptual view of this.

A

um Nih is funding basically an initiative because they believe untapped potential to contribute new expertise. New recruitment strategies, cutting edge science to ai and ml, can come from underrepresented communities, and so we're really trying to increase participation and engagement with these underrepresented communities to produce future researchers and also aid in scaling up and upscaling upskilling. Excuse me, current researchers uh to use ai and ml to um you know, model uh applications that pretty much um are um health disparities for for those under representing populations.

A

Excuse me some reason. uh Oh there it goes. Okay, all right so um so uh so there's a big coordinating center um and the coordinating center has uh uh four cores involved, and so again I I'm the pi of the data science training core uh and it's kind of a weird thing. This this was funded by what is called an ota.

A

um This is my first time actually being involved in the ota mechanism, but it's a different. It's not a grant. It's not a contract. It's it's! It's just something a little bit more flexible, uh but um when we went through this process it was very intense. You know it was like a one-week process. Where we had to submit you know applications.

A

Then we had to do a pitch, you know uh uh to nih and then what ended up happening is nih took what they thought were the you know some of the best uh uh proposals and they brought everybody together. You know, so there were multiple leadership proposals. They brought all them together and put them, uh and they said.

A

Okay, now, you guys go form a core and there were multiple infrastructure proposals and they said uh you guys go form the infrastructure core and uh the data science, uh the data research core as well, and then for for the data science training core. um We already had a consortium of universities that we went in, and so we did not get. uh uh We didn't have a forced marriage as as some of the other, uh some of the other groups.

A

Okay, so I won't talk through each one of these I'll definitely leave the slides for you guys. I want to really focus on the data science training core and just uh what the purpose of it is. Okay, and this was directly from the.

A

Excuse me all right, so um so the the data science training core our purpose is pretty much to uh identify training, needs and gaps um and develop uh training and workforce development resources, and we want to develop training around ai and ml uh and health disparities research, but we also are involving training for related competencies. So, as you see, there's a vast amount of competencies there from cloud computing distributed computing, but also uh even prerequisite type knowledge. So you know to get someone: that's a practitioner to do actual data science and ai and ml.

A

Maybe they might not have some of the mathematical prerequisites so they're actually saying well. How can we upskill these these these these people, uh if they want to engage in any form of of um practice using ai in in in ml?

A

um It's expected that you know whatever we create the engagements involve uh very hands-on, pedagogy and inquire and use real data. You know from real data sets from uh ehr data as well as public data. You can see the large team that we have here at howard. I won't go into the details, but we, you know a lot of interest across faculty from our hospital from our dental school, from our nursing, allied health and pharmacy, uh from arts and science, and and also from computer science and engineering.

A

And so we have like a you, know, an aiml center of excellence. We have a new uh uh center uh for applied data science, um you know and- and we have obviously phd bachelors to phds and computer science as well as a master's in applied data science. So we're well positioned to to to to to lead the effort, but we have a lot of partners too as well.

A

Our partners are medstar health and uh university of maryland and uh um universities um and and we we were working together prior to this for 15 years through an effort called guts which stands for the georgetown howard uh um center for clinical and translational science. We actually actually offer a a a certificate and a master's degree uh in in in in that in that area, but we also have partners with oak ridge.

A

National uh labs in the dc uh va medical center so again leveraging all the strengths we all came together for this initiative with three goals: one uh you know um wanted to identify training, needs and gaps uh and then classify them into near mid, mid and long term initiatives.

A

Second, we wanted to develop uh and expand uh and strengthen existing uh educational programs at the certificate undergraduate graduate and professional levels um and really focus on you know: training the a new generation of minority and underrepresented populations to become future researchers uh and third uh being able to coordinate all of these efforts across all the different cores and the stakeholders that are involved.

A

So um we we wanted to kind of look at how we can impact, who we can impact in the in the time scale that we can impact them so in the in in the one to two year. Well, let me let me scroll back the the initiative is 100 million dollars two year effort, you can say what are you going to get done into your efforts so um that that's the big question?

A

So um it's it's a two-year effort, but the the the uh the grant opportunity renews every two years um and what's going to happen, is um we're going to be opening up for new partners to come into the consortium. um You know over time, so we're like the first. I guess we're the first sets of partners that are coming on to their mission, um and so in the near term. We we want to make sure that we can target.

A

You know in a one to two year time frame, um early career faculty, faculty, research, faculty and scientists and postdocs, obviously, practitioners in the medical field in the midterm. We want to leverage existing programs, so we can expand the pool. You know get more uh students that are in computer science and engineering, mathematics to say: hey. There is a career opportunity for you to take your skill set and apply to uh health disparities in that particular field.

A

So looking at bs masters phd programs, um but also how can we take a non-engineering program um and then and integrate um a curriculum in in those programs uh as well uh to use ai and um machine learning uh in the long term? We want to look at how we can look at existing workforce and uh provide opportunities to upskill or re-skill those.

A

We also want to look at at the undergraduate level. How can we get more students to become to go into the field of computer science and data science? That don't have that background, because traditionally those those fields are like elitist?

A

You know you have to have the prerequisites and knowledge and most underrepresented groups, don't they they haven't been exposed to computer science early as they're they're, they're they're, counting apart. So so you know how do we, you know, get them into a program? How do we bridge them into um into graduate programs uh uh uh where they can sit?

A

You know, uh you know side by side well in a traditional program with with students um that may have that background and and and they don't experience uh or they, you know, we lessen the imposter syndrome that may occur because of the bridge and that that we do and then obviously we want to look at national programs that that are offered at the high school level. So we provide uh various training pathways from high school.

A

You know our our goal is to provide pathways from high school all the way up to uh practitioners, faculty, postdocs, research, practitioners um and so again, this right here kind of explains in in I guess in detail um what what we're offering um um um to to to to the community and that community is uh partners. So our partners are the people that you know the organizations and people that are involved uh with the developing the the the center the coordinating center, but we have stakeholders, so our stakeholders are students.

A

Our faculty uh practitioners work workforce people in the workforce, um and so we want to be able to provide these these pathways for for our stakeholders um and the model that we chosen is through grants uh uh and training opportunities and so for high school students. We want to provide a way for stakeholders to be able to.

A

You know, get a grant to develop modules, to support to support existing national programs or or local programs that they may have um in in in their area, and hopefully it it sees or inspires students to go on and to uh get it. You know go on to gra undergraduate degree programs in data science or computer science, but we also want to look at workforce development and also students that may be an undergrad um that may you know, want to go into a career in data science, and so how do we bridge them?

A

And so the idea is to come up. We we have a um at um at howard, university um and through our consortium um um we have a four year. uh Excuse me a four-course sequence um as well as some wrap-around support that allows.

A

Students to um you know, try out uh the the the the major you know um uh and their the tuition is subsidized, so it's like they kind of try before they buy and the idea uh as they're taking those four sequence courses we're integrating some content from the data science certificate program from google, as well as giving foundation and programming algorithms data and so forth, um and and and again the mentoring and wrap around support so that, if they so choose to uh go, uh hopefully go into a masters or or a master's degree program in data science or cs to transition is, is a little bit um less of a burden.

A

um But if they don't so choose, let's say they you know for for some reason you know um life happens, and you know they have to do some other things. um They at least get a google certificate at the at at the end of the day um for existing students at the bachelor's and master's and phd levels. What we hope to do is provide opportunities for them to you know, take courses take courses that are developed through this process. So what we're, assuming is that faculty members, researchers and so forth?

A

Could you know apply to many grants to uh you know develop courses. We also want to look at new career faculty and allow them the opportunity to be able to buy out of a course with the assumption that they would create a course and and offer the course on their on their campuses camp campuses. um We also want to offer uh grant opportunities for people to develop modules again to go into existing courses. um uh Primarily you know non-computer science or engineering focus.

A

We also want to offer uh opportunities for professional development. So um you know a researcher or practitioner or faculty. Member or organization could offer workshops, seminars, boot camps, executive training or conferences for those opportunities, and then we wanted to have some opportunity for students masters and phd students to get research training practicums.

A

So um we wanted to allow them to apply to programs where we would provide assistance throughout the entire year and place them in one of our partners, uh facilities working with a mentor and then obviously outreach um outreach at the undergraduate level, but also at the k-12 level, again developing modules or developing entire programs to again to increase the uh to expand the population.

A

And we hope that through all these different programs, we hope to expand the pool. One thing that I do want to talk about is that all of the curriculum that's being developed is is, is utilize real data um from ehr or public data, and um the laboratories and projects should involve hands-on pedagogy.

A

We hope that um one of our main goals is to make sure that all of the uh everything we do is inclusive, um and, and so we wanted to make sure that the content and the curriculums that are that are being developed are socially and cultural use. Socially and culturally uh relevant pedagogy, and they also involve social and environmental determinants of health focus on structural, racial inequities, race, gender and ethnicity, as constructs for uh health disparities.

A

So those are one of those some of some of the requirements that we're gonna uh put in in in in the uh the uh the funding announcements, as as we start to disseminate them. The last thing uh everything is kind of um what we. What we hope is.

A

We have to look at sustainability, and so we um we were um charged with um developing uh a a a a help desk to support all the initiatives, and so what we see you know we envisioned is that you know every all of the content um that that is created becomes uh open uh source through some, let's say a creative commons license, um and and and it's uh can such that anyone can take it and use it remix it um for for their own purposes, but it.

A

uh But we also wanted to have wrap around help desk support so that, if, if anyone has questions, um they can easily um engage in what we call a community of practice uh so that you know um this could uh live and grow uh on its own. Assuming the funding goes away at some point, we don't know when that may be, but uh right now, they're saying it's until congress says they don't want to do it anymore. They don't they they'll continue to fund it.

A

So just um two more two more slides just so you can get an idea of how um how you potentially could gauge assuming that you're one of one of these potential stakeholders so for students. Again, as I said before you, they could take a course attend workshops, seminars, conferences that are offered through the the consortium they can apply for training practicums.

A

They can get support through the help desk. They can apply to the bridge program if they're, a student that may be from a non-cs non-engineering math background to potentially go on to graduate programs, and they can support any kind of outreach initiatives because they will be looking for student assistance.

A

Faculty members can apply for grants, collaborate with other members of the acc to hopefully form uh you know, uh training opportunities and or new courses um can apply for grants and serve as mentors for students that have applied for practicums.

A

They too can also get support and engage in community practice, and they can also apply for grants, for course, buyout uh and also provide opportunities for outreach as well. Practitioners, postdocs can do a similar thing um as faculty um and then uh workforce.

A

Anyone in the workforce could take class, take take any of the existing courses attend workshops, um but they can also apply to uh uh uh bridge programs. So we we, we think that there's going to be a lot here for workforce. You know um you got boot camps and so forth. So we think a lot of you know. We think we'll have a lot of uh applications for people that want to uh stand up these kinds of these kind of things for the workforce um external partners.

A

um Again um we do have commitments from from south uh uh the south uh big data hub and and various other members. This is just the beginning. This is just the beginning of some of the partners. We we're getting a lot of requests for or for additional partners to come online, um but we see partners playing um various roles and I just kind of kind of outlined a lot of them here um with industry.

A

You know again, we want to alleviate as much friction as possible, so we've got some commitments for licenses and credits uh towards computational resources, um but you know a lot of industry has trickled um as well, um so we want to make that available, to uh course, developers who apply to grants um to to create these courses uh for conferences and workshops. Again we have a similar agreement uh for practicums. um You know we we, you know our process is a mini grant model.

A

As I said before, work where you know, uh practitioners or or partnered organizations could apply for grant opportunities, and so what we're gonna need, reviewers that will come from the community, um so obviously partners could serve as reviewers uh and then and and they provide technical support for the resources that they are providing um organizations like uh south big data hub, you know, definitely um could help out with you know: curriculum um um access to members um for disseminating uh uh funding announcements, identifying resources. uh They can also apply to grants.

A

So I could see conferences and things like that that they could apply for again, assuming that it's catered towards um the same goals as the eight as the aim ahead initiative um and then uh again, medical resource or medical and health organizations and laboratories. um We've gotten a lot of commitment for again data compute power as well.