Talk Title "Montana Partnership for Broadening Career Pathways for Mathematics and Statistics PhD Students"

Presentation given by Dr. Lisa Davis at ICIAM (International Congress on Industrial and Applied Mathematics) on July 19, 2019

MT PEAKS Co-PIs and Team:

John Borkowski - Statistics

Karlene Hoo - Graduate School

Stephen Sofie -  M&IE  and  Materials  Science

Doug Cairns - M&IE  and  Materials  Science

Roberta Amendola - M&IE and Materials Science

Doreen Brown - CHEM & BCHM and Materials Science (Evaluator)

(slide #1)

MT PEAKS (MonTana Partnership for Enriching mAthematical Knowledge and Statistical skills)

Partnership between Math Sciences and the Materials Science (MTSI) PhD program

(Original Slide showed pictures of the following individuals):

Lisa Davis

Mathematics faculty

Math Grad Program

Karlene Hoo

Dean of Graduate School

ChBE, MTSI faculty

Doug Cairns

M&IE, MTSI faculty

Doreen Brown

MTSI Program Coordinator

Assessment

John Borkowski

Statistics faculty

Stat Grad Program

Stephen Sofie

MTSI Assistant Director,

M&IE, MTSI faculty

Roberta Amendola

M&IE,  MTSI faculty

(slide #2)

Thanks for support from

Rob Walker

MTSI Director,

Chemistry faculty,

Letter of Support

MTSI Lab Coordination Participant Projects, Courses

Beth Burroughs

Math. Sciences Head,

Letter of Support

Course Development

Doug Cairns

M&IE, MTSI faculty

Boeing Letter of Support

Aaron Goodwin

Formulation Development,

Capsugel/Lonza,

Letter of Support

Hosted Math Intern

NSF - EDT

PD Matt Douglass

NSF DMS-1748883

(slide #3)

AMS Employment Data and History...let's do the numbers

Table: AMS - American Mathematical Society Annual Survey Data, (EENDR) with Response Rate of 50%

∗BIG = Business, Industry, Government

"Workforce Development in Mathematical Sciences" program with start date of August 1, 2018.

(slide #4)

Enriched Doctoral Training (EDT) in Mathematical Sciences

Funded through NSF-DMS and MSF-DGE under Program Manager Mathew Douglass

Goals:

• Strengthen the nation’s scientific competitiveness by increasing the number of well-prepared U.S. citizens, nationals, and permanent residents  who pursue  careers in the mathematical sciences and in other professions in which expertise in the mathematical sciences plays an increasingly important role.
• Prepare participants for a broader range of career paths than has been traditional in U.S. math/stat doctoral training. (#1 for us)
• Prepare PhD students to recognize and find solutions to math/stat challenges arising in other fields and in areas outside today’s academic setting.
• Involving PhD students in research activities supplementary to the dissertation theme, but we have to minimize any potential increase in the students’ time to degree.
• Encourage connections between math/stat academic departments and other units within and outside of the university, including those in BIG.

(slide #5)

About the PEAKS Program

1. One year program - Math and Stat students apply, participants are matched with MTSI faculty labs based on interest
2. MTSI Lab Rotations: Two 8-week rotation(s)
3. MTSI Lab Exposures: 3 hours/week in a lab shadowing a MTSI graduate student, attending weekly lab meetings, and interacting with both MTSI students and faculty.
4. Interdisciplinary Research Project, BIG-inspired when possible
5. BIG Internship 
6. Professional Development Workshops (GS), Coursework MTSI Intro course

Ongoing Activities:

Assessment: Doreen Brown is evaluating the program during every activity with Pre- and Post- Interviews and Rubrics

(slide #6) 

Overview of Program and Participants

3 Types of Student Participants

PEAKS Fellows - 3 per FY

• MathStat PhD Candidates

• Two 8-week MTSI Lab Rotations
• GRA appointment
• Small $ for lab supplies

PEAKS Affiliates - 8 per FY

• MathStat MS or PhD students
• One 8-week MTSI Lab Exposure
• Small $ for lab supplies

MTSI Lab Mentors - 2-4 per FY

• MTSI PhD Candidates
• Mentors Fellows and Affiliates in labs
• small $ for each 8-week session

(slide #7)

Moses Obiri - Stat PhD

John Borkowski (Stat), Stephen Sofie (MTSI) and Stephen Heywood (Lab Mentor)

Develop optimal space filling design for algorithms for the combined mixture/process variable experiment

• Investigate the effect of vanadium (V), molybdenum (MO), and strontium(S) mixture ratios on relative density, phase purity and electrical conductivity of a solid fuel cell.
• Investigate the effect of citric acid, ethylene glycol, and temperature on relative density and electrical conductivity of a solid fuel cell.
• Sintering Methods at 1400◦C  -  form  a  solid  mass of material by heating without melting to the point of liquefaction.
• Example of Three-Component Mixture Experimental Design Problem
• Space-filling designs
• Nuances and complexities of conducting experiments

(slide #8)

Catherine Potts - Math PhD

Dominique Zosso (Math), Anja Kunze (MTSI) and Derek Judge (Lab Mentor)

Calcium imaging of neurons to find network connections

• Image processing and machine learning techniques
• A representative dictionary via Archetypal Analysis
• Develop Archetypal Analysis metrics applied to calcium imaging
• Compare with current calcium imaging metrics
•  Get neurons from a rat brain (breaking communication pathways)
• Add Fluo-4 dye wait ≈ 60 minutes
• Record under UV-stimulation
• Record calcium image for comparison
• Add nanoparticles wait ≈ 6 hours
• Record under UV-stimulation

(slide #9)

Steve Walsh - Stat PhD

John Borkowski (Stat), Roberta Amendola (MTSI) and Madisen McCleary (Lab Mentor)

Effects of Al₂TiO₅(Aluminum Titanate) doping on the strength of NiO YSZ (yttrium stabilized nickel oxide) ceramic materials used in fuel cells

• Developing a reliability model to identify mixture of Weibull distributions in strength data
• Discrimination/classification of specimens
• Developed measure to confirm when one specimen set may contain multiple fracture distributions 
• Sample Prep by tape casting Strength Testing and Data Collection
• Electron microscopy - characterize sample morphology & fractographic mechanisms (intergranular and transgranular)
• Multiple Weibull Distributions
• Reduce the # of specimens needed to characterize fracture

(slide #10)

NSF's Intended Benefits and PEAKS Outcomes

General Outcomes:

• Broadened graduate training format in MathStat
• Graduates actively select non-academic careers
• More diverse workforce with STEM skills established

Grad Students:

• Enhanced understanding of their research areas in a broader context
• Academic careers: Math/Stat inspiration from problems in other disciplines, links between academics and BIG/non-profit realm; Advise future students on broad career paths
• Outside academics: Better preparation for career paths in BIG/non-profit

Faculty and Academic Units:

• Enhanced recruitment of and placement of students
• Enhanced ties to other disciplinary units and non-academic partners

Academic Community:

• Suite of pilot projects for grad programs nationwide, especially those who have limited industry contacts and geographic challenges for industry diversity

(Slide #11)

Outcomes

Outcomes:

• Paper recently accepted in Materialia by Dr. Roberta Amendola, Dr. Madisen McCleary and Steve Walsh
• Minisymposium talk application for BMES 2019 (Biomedical Engineering Society).  Archetypal Analysis for Nanoparticle-mediated Calcium Signaling.
• 3 for 3! Summer internships at national labs

Catherine Potts - Applied Machine Learning Research Summer School at LANL

Steve Walsh - Applied Statistics and Computational Modeling Group at PNNL

Moses Obiri - Group at PNNL

• Presenting results at MT PEAKS Annual symposium (September 26-27) joint with the MTSI annual symposium
• Flyer circulated and program advertised at the Grad Recruitment Weekend in Feb. 2019. Potts and Walsh gave 10min. presentations to the students.
• MT PEAKS Website

(Slide #12)

Lessons Learned

• Improved communication of procedures and expectations to students
• Matching students (interests) with MTSI labs is crucial!!
• Lab Rotations: More flexibility in time durations
  • Rotation was slow to start and ended too quickly - it was just not enough time
  • How much time do we need to spend in the lab? - How is the pairing of faculty with students going to happen? - What (safety/instrument) trainings were expected of them?
  • MTSI grad student was not as enthusiastic about letting him get the true “hands-on training" regarding the use of instruments despite believing he was
•  Lab Exposures: more hands on training, especially w/instruments less observation
  • Affiliates were surprised as to the amount of time and expertise it took to design  and execute a lab
  • Affiliate spoke positively of the MTSI students & lab personnel - exposures were professional and friendly
  • Adequate resources to conceptualize the difficulty and complexities of planning a MTSI experiment

(Slide #13)

Thanks for support

Faculty Mentors 

Assessment, Feedback and Overall Reality Checks!

This work supported by NSF EDT program through DMS-1748883 (PD Matt Douglass)

(Slide #14)