Karan is a double major in physics and computer science at the Georgia Institute of Technology. He loves simulating physical systems on computers, and plans on going to grad school for computational science and engineering. He also likes studying history, psychology and economics.
He currently works in a quantum computation lab and a high-energy astrophysics lab at Georgia Tech. He is also interested in online education and has been involved in the creation of different MOOCs. He loves making things: code, electronics or imaginary constructs. In his spare time, he loves reading (pretty much anything), playing pool, watching movies and biking.
Project: Classifying Cellular Automata Using Machine Learning
While cellular automata can be classified manually for small spaces (such as 256 elementary CA), we need more efficient and automatic methods to classify them over bigger spaces. I used unsupervised classifying methods to do so. To make it more effective, I came up with preprocessing functions and feature sets for the classifiers. The flowchart below shows how the algorithm works.
Favorite 3-Color 2D Totalistic Cellular Automaton