Machine learning for high-throughput phenotyping
Drone alert! In this collaborative project we used drone images and a simple machine learning algorithm to capture variations in flower opening time in a lettuce population. Preprint is now available!
Hello there, my friend!
Welcome to my humble personal website. I am a 4th year PhD student at University of California, Davis. A member of the marvelous Michelmore Lab, I am a plant geneticist by training, and a crazy data lady at heart. My research interest is to extract biologically meaningful signals from large datasets generated from high-throughput platforms. These include remote-sensing phenotyping data, transcriptomic data, and genomic sequence data. I try using a diverse set of tools in my analyses, ranging from machine learning to Bayesian inference to various linear models.
Before coming to Davis, I studied Agricultural Sciences at Cornell University and graduated Summa Cum Laude in 2015. It was in the #gorges embrace of Ithaca, NY that I developed an interest in plant genetics. Check out my CV for a detailed chronicle of my professional life.
I love all things genetics, genomics, and statistics.
Biologist and Statistician
I have the "double identity" of being a biologist as well as a statistician. By that I mean I am interested in understanding the biological mechanisms underlying various life phenomena, as well as in developing quantitative tools for the discovery of these mechanisms. Here are some of the projects that I am recently involved in. I hope they pique your interest in biostatistics!
Bayesian statistics
In the R club organized by Prof. Julin Maloof at UCD Plant Biology, we are learning Bayesian statistics using Statistical Rethinking by Richard McElreath.
Causal inference
You have a p<0.05, and now what? I was recently introduced to the causal inference framework from the UCD course STA207. My group's term project showcases an example of using propensity score for making causal inference in observational studies.
Machine Learning for forecasting
The UCD course STA208 provided a comprehensive overview of the mathematical bases and Python sklearn applications of many methods that fall under the umbrella term of "machine learning". My group's term project for this class used a selection of them, namely, k-means clustering, kernel PCA, deep neural net, and random forest regression, to integrate information from two independent sources to predict the number of bike trips happening between each pair of stations in the Washington DC's Captical BikeShare network.
Things I Enjoy
It's hard to believe but life goes beyond biology and statistics! When I'm out of my office, I keep busy with hobbies.
Contact Me
If you see a project that you have questions about or want to collaborate on, please don't hesitate to reach out! Like Helen Keller once said, "alone we can do so little; together we can do so much."