Alex Roman - Doctoral Researcher - University of Florida | LinkedIn
$$ \text{take action}\ a = \argmax_{a\,\in\, \text{actions}}\, (\, \mathbb{E}[\, \mathscr{U}| a]\,) $$
where $\mathscr{U}$= my utility function. This implies three key questions: What is $\mathscr{U}$? What is the set of a’s, and how do we quantify the distribution in order to take $\mathbb{E}$?
Hi, I’m Alex!
Most recently I was teaching Machine Learning as an Adjunct Professor at New College of Florida.
I’m a doctoral researcher at UF, and I’m interested in quite a few different things.
While doing my undergrad and my PhD I have been on a journey from the theoretical side of theoretical physics writing my undergraduate thesis on “Black Hole Thermodynamics,” to the state of the art in Machine Learning and AI, which I’ve been applying to problems in Particle Physics, and Exoplanet Astronomy.
This website is a living document which is supposed to reflect all my different interests, research and otherwise. I am not done writing all the different summaries and explainers
<aside> 💪 Cooperates in the prisoners dilemma!
</aside>
One boxes on Newcombs problem!
Getting Started in AI Research
The first thing you should know about Quantum Computing is that there are two very different types: Circuit Based Quantum Computing and Adiabatic Quantum Computing. The former kind is the one that gets more attention in the media, but my work focuses more on the latter. One reason why we should care about Adiabatic QC is that the technology already exists and is capable of being used to solve hard real world optimization problems now. What is now needed is understanding of how to encode problems in the Hamiltonians on which these systems are based.
Adiabatic Quantum Computing Explained