Postdoctoral Associate
CurrentResearch machine learning.o Collaborate with Dr. Bo Deng on developing methods for completely training a classification neural network to 100% accuracy. Previous methods do not attain 100% accuracy for predicting train and tests data.▪Run MATLAB and Python scripts on a supercomputer.▪Access supercomputer via a Linux terminal.▪Submit scripts using Slurm—an open-source job scheduler.oAdvise undergraduates.▪Guide students through reading a research paper.▪Teach students how to construct a neural network from scratch.Collaborate on a project studying dimension reduction in complex projective space.o Program in Macaulay2, a software for computational algebraic geometry and commutative algebra.Leverage my broad expertise to teach an experiential Data Science capstone course which utilizes machine learning, financial mathematics, and statistics.o Partnered with the Lancaster County Assessor’s Office to teach an interdisciplinary course where students conducted a mathematical analysis using real data.▪Group 1 compared a neural network model to a simple linear regression model.▪Group 2 built a decision tree to determine the best neighborhoods to buy an investment property in Lincoln, NE.▪Group 3 used hypothesis tests to determine what features were most predictive of a home’s price.▪Group 4 used housing sales data to approximate a supply curve and determine the average price of the Lincoln housing market. They leveraged this to predict a demand curve and measure volatility.Develop curriculum.o Served on curriculum development committee for second year math courses.o Created programming labs for UNL’s linear algebra course that introduces students to linear regression and computer vision.Supervise multiple teams.o Lead the trigonometry instructional team.o Chair the Math Day committee. This involves leading a team of graduate students, front office staff, and other faculty.Teach graduate and undergraduate courses.
