This semester I am assisting a team of 4 undergraduate students including an exchange student working with their research project. This project involves implementing a numerical method for very rapidly and concisely approximating a function of many variables based on only a few (random) function evaluations. This problem is very challenging for arbitrary smooth functions due to the rapid blow up of the complexity of such functions with the dimension of their domain (i.e., the number of variables on which they depend), a phenomenon generally known as the ``curse of dimensionality". As part of this project the student participants are trying to undo the curse by using a collection of techniques from compressive sensing, probability, and computer science in order to implement an ultra-fast method for accurately approximating, integrating and interpolating a large class of functions which are simply too hideously high-dimensional to handle in any other way.
| Semester | Course | Role |
|---|---|---|
| Spring 2016 | Survey of Calculus I (MTH-124) | Instructor |
| Fall 2015 | Calculus II (MTH-133) | Teaching Assistant |
| Spring 2015 | Survey of Calculus I (MTH-124) | Instructor |
| Fall 2014 | Calculus II (MTH-133) | Teaching Assistant |
| Summer 2013 - Spring 2014 | Survey of Calculus I (MTH-124) | Instructor |
| Fall 2012 - Spring 2013 | Calculus III (MTH-234) | Teaching Assistant |
On March 4th 2017, I assisted activity titled "Group Testing : How to find out what's important in life" for middle school students as part of "Girls Math and Science Day" at Michigan State University. The students learned general "binary search" techniques that were used to find soldiers infected by syphilius in WWII. They performed a safe experimental implementation of a group testing procedure. Each group of students was given 4 to 8 "mystery cups'' each containing one of two possible types of powder (either bleached flour, or flour mixed with baking soda). The group's job was to identify which mystery cups contained flour mixed with baking soda. These cups could be tested for the presence of baking soda by mixing vinegar with samples of the powder from various subsets of the mystery cups in a separate "test cup". Each group deduced from the experiments the fewest number of test cups to correctly identify the one (or two) mystery cups containing baking soda.