1. Best route for grad school? and 2. How much do publications matter?


New Member
Apologies if I come off as annoying. I'm currently a sophomore at a non-target state school majoring in Electrical Engineering. My GPA is a little above 3.00 for my major, and it's 3.1 cumulative. A little about my background:

* I conduct research in signal processing and deep learning, where I have to train and test deep neural networks to validate the results of my team's proposed unsupervised algorithm for brain network time-series clustering. The goal is to get published this fall by IEEE Signal Processing Society.

* I code in Python, R, C++, and Matlab, and I do independent research to the point where I've proposed a new optimization algorithm that's being reviewed by one of my professors

* I have numerous projects to show for my skills in machine learning, deep learning, and time-series analysis, including the testing results for the neural nets used in my research (I designed the architectures)

*Relevant coursework: Calc 1-3, ODEs, Linear Algebra, Applied Probability, Physics 1-3, Signals and Systems

*Currently independently learning: Stochastic calculus, stochastic processes, measure theory

All in all, I'm looking for advice on how to approach applying to grad school. My GPA isn't very strong, and I know my skills and projects won't make up for it. So I'm wondering if I should aim for more publications. I love research and I want to conduct more in artificial intelligence, signal processing, and even physics (statistical thermodynamics) for their applications in finance. For example, Brownian motion in the Heston Model and the Ornstein-Uhlenbeck process. What should I highlight? What is relevant? And do I stand a chance at the Top 15 or so CS PhD programs? Or should I study something else in grad school like statistics?
Last edited: