UROP Research Mentor Project Submission Portal: Submission #571

Submission information
Submission Number: 571
Submission ID: 9126
Submission UUID: 4eb264cb-bd16-4228-b65b-cbdfe56730aa

Created: Thu, 08/24/2023 - 06:08 PM
Completed: Thu, 08/24/2023 - 06:08 PM
Changed: Mon, 08/28/2023 - 09:39 AM

Remote IP address: 217.180.196.52
Submitted by: Anonymous
Language: English

Is draft: No

Research Mentor Information

Alexander Reznikov
{Empty}
None; just write "Hello Alexander"
areznikov@fsu.edu
Faculty
Arts and Sciences
Mathematics
{Empty}

Additional Research Mentor(s)

{Empty}
{Empty}
{Empty}
{Empty}
{Empty}
{Empty}
{Empty}
{Empty}

Overall Project Details

Gradient descent on the unit sphere
Calculus, potential energy, optimization, gradient descent
Yes
1
Preferably mathematics or computer science, but open to all majors
On FSU Main Campus
{Empty}
Partially Remote
7
Flexible schedule (Combination of business and outside of business. TBD between student and research mentor.)
Imagine putting four repelling particles on the sphere. They will move around the sphere until they reach the "ground state": a position where the combined potential energy has a local minimum. If we are not too unlucky, they will actually stop at the global minimum: i.e., the regular pyramid.
The goal of this project is to figure out what does it mean to be unlucky and experimentally show that this is a very rare situation.

This is done using "gradient descent", which is technically taught in Calculus 3 (MAC 2313). We throw four random points on the sphere and move it, one little step at a time, in the direction where the energy decreases (a.k.a., in the direction of the gradient).
Collect specific data of initial random configurations that are "good" (i.e., end up at the global minimum) and "bad".

If time allows, do the same in higher dimensions (this is not strongly expected).
Required: Calculus sequence, some coding skills.
Recommended: Having C++ or Python (preferably) installed, some confidence in coding.

We will discuss the approach, and you can learn the specific tools/packages along the way, but I would strongly prefer you to be able to open Python on your own.
Passing a course without applying it makes the course useless in the long term because things get forgotten. I view this project as an excellent opportunity to combine your Calculus knowledge and coding skills to solve a real-world problem. The mentee will not only get a taste of real research environment, but also some understanding of how to find practical applications of what seemed to be "pure theory".
{Empty}

UROP Program Elements

Yes
Yes
Yes
Yes
{Empty}
2023
https://cre.fsu.edu/urop-research-mentor-project-submission-portal?token=BBXZZrZzus0bQWAt-2B47asynY290iZkq81K8oxF_0g