Machine Learning in Public Health

Project Title

Machine Learning in Public Health: Understanding the ongoing opioid epidemic in the U.S

Faculty Mentor(s)

Project Description

Project Description: 

I did research about Machine Learning in public health under Professor Adam Eck's guidance. The research was focused on making statistical and machine learning prediction models to better understand the ongoing opioid epidemic overdose in the United States at the county level. We used 34 predictive features in our models, and the key features were unemployment rates and death counts. The data we used in training, testing, and validating the performance of our models was unrestricted, meaning it was available on the Internet. 

Some models included linear regression, random forest, super vector machine, and decision trees. I helped build a Rshiny website app that allows users to visualize the performances of the models in predicting the overdose death rates (death per 100 000 people) and counts (physical number of deaths). The website app has a geographical map of the U.S. that illustrates the distribution of opioid overdose rates and counts.

 

Why is your research important?

The research is key to understanding the ongoing opioid overdose epidemic in the U.S. at the county level. The study involved building a Rshiny website that allows users to visualize counties with high predictions of opioid overdose deaths. Furthermore, the users can also understand the pivotal prediction features in our models through the website.
 

What does the process of doing your research look like?

My research project involved writing code on my computer, so I did most of the work outside the lab. I meet with my mentor twice a week to discuss my progress. I work collaboratively with my research partner and sometimes meet daily to share ideas on the project.
 

What knowledge has your research contributed to your field?

My research has improved other people's understanding of the opioid epidemic in the U.S. and hopefully, guides public policy and decision-making regarding opioid prescriptions.
 

In what ways have you showcased your research thus far?

I helped build a Rshiny website app that people can use to visualize the regions of the U.S where there are high predictions of opioid overdose deaths. In addition I have presented my research to an audience of 40+ people.

 

How did you get involved in research? What drove you to seek out research experiences in college?

I am deeply interested in machine learning, so this research helped me learn and do more hands-on activities.
 

What is your favorite aspect of the research process?

My favorite aspect of the research process was learning a new programming language named R. I also learned how to build a website app library in R called Rshiny.
 

How has working with your mentor impacted the development of your research project? How has it impacted you as a researcher?

My experience working with my mentor was great overall, and inspired me to seek for similar research opportunities in the future.
 

How has the research you’ve conducted contributed to your professional or academic development?  

Doing research has inspired me to learn more about Machine Learning. I have developed my analytical and programming skills through the research experience.
 

What advice would you give to a younger student wanting to get involved in research in your field?

My advice would be that doing research as an undergraduate student is a great opportunity to explore your interests. As for me, I enjoyed learning about Machine Learning in my research.