Kartikeya Sharma
Kartikeya Sharma earned his Goshen College B.A. with majors in Computer Science and Accounting in 2019. He worked for a few years as a Data Scientist with Veada. He is currently in graduate school. Here a a couple of his updates about his progress.
September 6, 2021: I will be starting my degree in Masters of Science in Computer Science at the University of Oregon this fall. I have also been awarded a Research Assistantship in the Center for Cyber Security and Privacy under Prof Jun Li, where I will be researching Online Social network Fraud and Attack Identification
May 17, 2022: My first year at the University of Oregon is almost coming to an end with final exams due in the second week of June. I would like to share my experience from this (spring) and the past (winter) term, and my plans for the summer.
First of all, let me tell you about the boring part which is my research (though I am absolutely enjoying it) 🙂 I am working on spam detection on Twitter with the help of Graph Neural Networks which is a Machine Learning technique used to learn from Graph data structures. Last term (Winter) I was able to reduce the training speed on the training dataset to seconds when trained on CPU. When compared my implementation to the benchmark implementation, it was about 717 times faster. This opens the door for real-time machine learning as well as economical machine learning as most methods currently use GPUs which are very expensive. Currently, I am building on my research from the winter term to make the graph training work on dynamic graphs instead of static ones.
Besides research, I have taken or currently taking several different classes such as Computer and Network Security, Machine Learning, Data Science, and Multi-Agent Systems. I am absolutely loving these classes and learning a lot from them. My favorite class this term is definitely Multi-Agent Systems where we are learning about topics such as Security games, Behavioral game theory, Extensive-form games, Coalitional Games, Auctions, and their application with Machine Learning. Also, one thing which I like about taking classes at UofO is that they are very small in size just like it was at Goshen.