Meghan Lu

Meghan Lu
Meghan Lu

Meghan Lu

Year in School
Junior

Major

Resume

Year of Participation in STARS

  • Fall 2022

Research Interests
Artificial Intelligence Data and Information Systems

Research Mentor
Prof. Kris Hauser

Research/Engagement Experience
This will be my first time conducting research, and I am excited be able to learn about the research process and groundbreaking work being done on this very campus! In terms of engagement, I currently serve as the Social Co-Chair of Women in Computer Science (WCS) with the goal of creating a welcoming and inclusive computer science community on campus. I am also the Campus Co-Director of The Percentage Project, a nationwide social media campaign spreading statistics and raising awareness for the experiences of underrepresented minorities in computing. Previously, I have held chair positions for Society of Women Engineers (SWE) and Alpha Omega Epsilon Engineering Sorority.

Interests
Diversity in tech, mindfulness, fashion, and trying new foods!

Project Title
Auto Model Predictive Control (MPC)

How did I get interested in Computer Science?
I like math and problem solving, so I enjoy coding in general. However, the main pull of computer science is its potential to create new solutions in almost any area. Computer science has completely changed the world in a relatively short period of time, and the innovation and growth of the field is incredible.

What social interests matter to me?
Underrepresented minorities have many more barriers to access, and those that are already in the field face higher rates of imposter syndrome and will often undervalue themselves which only reinforces the gender gap. It is so important to create a space where underrepresented minorities feel like they belong in computing and are able to see successful role models. Though there has been a lot of work done in this sphere, it is still as necessary as ever. I am also very passionate about gender equality issues, including abortion rights.

What is my most impactful college experience?
Joining organizations like Women In Computer Science (WCS) and Society of Women Engineers (SWE) and taking on responsibilities that allowed me to ideate and bring events to life made me realize that I could have as big of an impact as I wanted to have, even at a school as large as UIUC. I have also been able to meet so many amazing and inspiring people through joining these and other student organizations!

These are a few of my favorite things!
Good weather, curating niche Spotify playlists, studying at Grainger Library with friends, getting to know a new city, art museums, closing my Apple Watch fitness rings, and rice pudding.

Research Description
Model predictive control (MPC) is a common method for robot motion control which uses a model of system dynamics to compute a system trajectory according to some cost function and replans as new information becomes available. The performance of MPC depends significantly on the accuracy of the system ID model, but designing a model for a novel system can be quite challenging and requires careful selection of model class and hyperparameters. In order to automate this process, the AutoMPC package searches for the best model for a system and tunes its hyperparameters. The Gaussian Processes (GP) model is the only non-parametric model supported by AutoMPC and outputs uncertainty estimates on its predictions which allows the model to be more interpretable. The GP model depends heavily on its kernel function hyperparameter, which compares the correlation of two data points in the state space. At this point, AutoMPC only implements a single option of kernel function for the GP model. This project focuses on implementing additional kernel options and evaluating their impact on modeling error and MPC performance.

Biography
Meghan Lu is a junior studying Statistics & Computer Science with a minor in Mathematics, originally from the Chicago suburbs. After her past summer internship working on autonomous navigation for robotics, her interest in robotics grew, and she will be researching AutoML for robotics controllers under Professor Kris Hauser starting this semester. Increasing diversity and inclusion in computing is very important to her, and she is heavily involved with Women in Computer Science (WCS) as a Social Chair and The Percentage Project as Campus Co-Director. She has also been involved with Taiwanese American Student Club (TASC), iRobotics, Society of Women Engineers (SWE) and other student organizations on campus!