Purdue is a research-led institute and I am very passionate on mentoring both graduate and undergraduate students to conduct research. Both university and Krannert have provided various scholarship opportunities to help students gain some early research experiences. I also use research grants to sponsor some students. Krannert's vision is to train analytics leaders in business world. Our research projects, either as an internship project or as independent studies, allow students across the campus to access the advanced topics under our supervision and experience the multi-disciplinary research environment.
UNDERGRADUATE RESEARCH PROJECTS
Undergraduate research can be funded by Office of Undergraduate Research (OUR) Scholarship or Discovery Park Undergraduate Research Internship (DURI) without course credit, or unfunded with independent study credit.
2022 Fall OUR Project: Supply Chain and Manufacturing Operational Excellence for Sustainable Business Models
Team: Joyce Zou (Krannert, Senior)
2021 Fall OUR Project: Optimizing Distribution and Allocation Strategies for Covid-19 Vaccines
Team: Hui Zeng (Sc&Info&Analytics, Senior), Congshi Pu (Sc&Info&Analytics, Senior)
2021 Spring OUR Project: Managing Supply Chain Disruption Risk for Coronavirus Outbreak
Team: Hui Zeng (Sc&Info&Analytics, Senior), Quan Wang (Economics Honor, Senior), Shuoyao Li (Sc&Info&Analytics, Senior)
2020 Fall OUR Scholarship: Managing Medical Supply Chain Disruption amid Covid-19
Team: Hui Zeng (Sc&Info&Analytics, Senior), Quan Wang (Economics Honor, Senior)
Abstract: The COVID-19 pandemic is threatening human lives and the world economy. Personal protective equipment (PPE) became an essential medical resource. The shortage of PPE was often reported by the media and it was perceived as supply chain disruption. U.S. manufacturers are blamed for relying on offshoring significantly and asked to move their supply chain backshore. Is it true that there was a significant disruption in the PPE supply chain during the pandemic? Is the global supply chain for PPE resilient for the U.S.? What’s the profile of the current global PPE suppliers? How to build a more transparent global PPE supply chain? What are the implications for policymakers to manage the supply chain disruption risks? We collect the PPE trade data from various sources, which allows us to review the facts in the global PPE trade before and after COVID-19. We next perform time series analysis to analyze the trend of the global PPE trade and difference-in-difference model to analyze the effect of COVID-19 on the global PPE trade. Counter to the general perception, we found there was no significant disruption in the global PPE supply chain; it is rather flexible and resilient. We use Python programming to construct a PPE global supply chain map and identify the number of suppliers with trade flows between countries indicating production capacity, which provides a highly transparent global supply chain map for decision-makers. Our study shed new light on the current debate about PPE disruption and U.S. manufacture backshoring.
Presentation on 2020 Fall Purdue Undergraduate Research Expo
2020 Fall: Long-term Care and End-of-Life Care (DURI and Smart Health Lab Internship)
Team: Sophie Qi (UPenn, Sophomore) and Evelyn Cai (Purdue, Non-degree freshman)
2020 Spring: Predictive Analytics for Coronary Heart Disease
Team: Megha Reddy (Public Health, Senior)
Abstract: Predictive modeling was done on a dataset acquired from the Framingham Heart Study using statistical methods like principal component analysis and logistic regression. This was used to identify high-risk cardiovascular disease (CVD) groups. Using this, my research makes recommendations directed towards managing these high-risk CVD groups during the COVID-19 pandemic.
The identified high-risk CVD groups included (1) older adults (2) males (3) smokers (4) hypertensives (5) pre-diabetics and diabetics. The odds of getting heart disease was calculated for each key risk factor and increased with increasing age, number of cigarettes smoked, and stages of medically diagnosed hypertension and glucose levels. These groups were also at increased risk of morbidity and mortality from COVID-19 and comprise the majority of those hospitalized for the virus. Recommendations were made to manage these high-risk groups during COVID-19 in the areas of (1) immediate isolation (2) reorganization of the medical task force and (3) medical management.
While there is an abundance of knowledge and data identifying high-risk CVD groups, we lack policies built around efficient population health management of these groups. There is a need to use big data to direct policies that recognize, prevent, as well as manage chronic conditions in these groups. This project makes risk-specific recommendations to manage adverse outcomes due to COVID-19 in these groups that are actionable through local public health departments and governments. This can potentially help curb poor health outcomes and lower the hospital burden posed by these groups, especially in times of health crisis.
2020 Spring Purdue Undergraduate Research Expo (3rd place under interdisciplinary posters).
2020 Spring: DURI Scholarship Project: Predicting Postoperative Deliverium after Intracranial Surgery
Team: Juliet Aygun (Mathematics, Sophomore), Alaina Bartfeld (Krannert, Sophomore), Sahana Rayan (CS and Applied Stat, Sophomore)
Abstract: Delirium has a high morbidity rate and is common; around ten percent of older, hospitalized patients have delirium, and fifteen to fifty percent of patients experience delirium during hospitalization. This puts delirium at the forefront of problems for which concern doctors and nurses look. The vast majority of journal articles about delirium focus on postoperative delirium (POD) in the Intensive Care Unit (ICU) ; however, none are specific to post-intracranial surgery. In this way, our research is distinct from others in this area. The purposes of this research project are to employ machine learning methods which accurately predict whether a post-intracranial surgery patient will be diagnosed with POD in the ICU, and identify the key predictors of POD. If POD could be predicted, many patients would experience a shorter hospital stay, less marginal complications, and a greater life expectancy. With our model, the onset of POD could ultimately be stopped. We first conducted detailed dimensional reduction on our dataset by employing factor analysis and elastic net classification to prevent overfitting of the model. After, we trained a neural networking model to predict POD. This model was eighty-five percent accurate, and we found the key predictors of POD are whether the patient had delirium when they were admitted into the ICU, the type of lesion they had (if any), and if their blood was carrying enough oxygen. This was supported by chi-square analysis, which proved all the variables are statistically significant at any alpha level. Our work has been recognized by different Purdue University research organizations. We are published in the 2020 edition of the Journal of Purdue Undergraduate Research (JPUR). In the journal, we have a smaller scale, abstract-like piece of writing that briefly summarizes our findings that will be published over the summer of 2020. In addition, we were selected for a 10-minute oral presentation in the Spring 2020 Undergraduate Research Conference that goes more in-depth about our work. In this presentation competition, our research project won third place out of all Krannert School of Management undergraduate researchers who also got selected for an oral presentation. We are grateful for the opportunities Dr. Zhan Pang has given to us through this project, and we are proud of the work we have produced with him.
Presentation in Spring 2020 Purdue Undergraduate Research Conference
Highlights of Some Excellent Students I Taught
