Arthur Lin presents in Wisconsin Falling Walls Lab

Arthur Lin, a graduate student in the Cersonsky Research Group in the Department of Chemical and Biological Engineering, presented this year at Wisconsin’s Falling Walls Lab in April. Named in honor of the fall of the Berlin Wall, Falling Walls Lab is an international competition where young innovators showcase their breakthroughs that have a positive impact on science and society. Presenters have three minutes to convince the judges of their research’s merit.

Taking a peak behind the “wall,” we learn more about Lin, his passion for machine learning and physical sciences, and the details of his presentation.

Can you tell us more about your presentation topic, Breaking the Wall of the Spherical Cow?

The title Breaking the Wall of the Spherical Cow alludes to the Spherical Cow metaphor used in physics, which basically states that while you can obtain some physical insights about a cow’s behavior/properties by approximating a cow to be a sphere, such an approximation loses grounding in physical reality. My spherical cow is the fact that we often don’t explicitly incorporate molecular geometry into molecular simulations. Within simulations, we often treat individual small molecules like water as spheres, or large polymer molecules as a bunch of spheres stringed together. Such approximations have been useful but also limited due to these oversimplified geometry assumptions. My work focuses on incorporating molecular geometry into molecular simulations, and using machine learning techniques to gain insights on how complicated molecular geometries influence the behaviors of molecules that we observe.

How do you feel the presentation went? What has this experience taught you?

The presentation went well but I was definitely nervous. Although I didn’t place top three, I was happy with my presentation. It was a two-and-a-half-minute presentation which was incredibly challenging, so it was a learning experience to understand how to frame the story within the very limited time I had. I had originally written an exact script, but didn’t have it memorized in time, so by the time of the actual presentation, I was presenting a bit off the cuff – in the future, I would probably just stick with a bullet point list of topics to talk about, because I found the script to be too stifling.  

How did your mechanical engineering studies help prepare you for your presentation? For beyond?

The engineering curriculum definitely places emphasis on communication and presentation, as many end-of-the-semester engineering classes contain presentations. My favorite presentations were in Professor Victor Zavala’s statistics class, where he restricted each individual to present their statistics projects within five minutes. I had a blast learning about a variety of topics within one hour, from the distribution of paper types in recycling mills, stock market predictions, and optimal experimental design. This competition was very reminiscent of these projects, though these presentations were even snappier.

What drew you to the Cersonsky Research Group?

Professor Rose Cersonsky is an excellent communicator and mentor, which were my top criteria when choosing a professor. I was always interested in the intersection of machine learning and physical sciences, and her strong background in developing powerful machine learning techniques that explicitly incorporate physics was a huge draw. Our lab is quite math heavy, so to see her being able to communicate these complex topics without getting lost in the math is something I admire and try to emulate, and I’ve been taking as many opportunities as I can to present my work, including via this competition!

What advice might you share with future UW engineering students?

Some of my advice as a second-year:

• The relationship between you and your advisor is definitely extremely important, so when you’re choosing one, make sure your personality type meshes well with the advisor.
• Listen to your advisor, especially when you’re just starting out!
• Perfection is the enemy of good (I really struggle with this).

Fun facts:

Flamingos or Badgers Badgers

Favorite Babcock Ice Cream flavor Cookies and cream, though I like to try a lot of different flavors!

Sweet Caroline or Jump Around? Sweet Caroline 100%

Favorite study spot? My office on the weekdays and some coffee shop like Barriques on the weekends.

Favorite place for lunch grub? On Fridays I treat myself to a Vietnamese sandwich from the Saigon Sandwich Food Truck. Wiscorrican food truck also has great Puerto-Rican food!