## AI and Machine Learning

Although it's first in the drop down menu because of the alphabet, this section would probably top a priority list, too. I've been deeply fascinated with autonomous machines since my freshman year in undergrad, during which I took a course called 'Mathematical Impossibilities' which revolved around the study of unprovability. However, the section that most strongly hooked my attention was the discussion of computability with regards to the Entscheidungsproblem (Decision Problem). In particular, I found Turing Machines to be absolutely fascinating- I made my first foray into programming (in the traditional sense, anyway) in order to build a simulator so that I could more easily experiment with them.

Later on, as I discovered electronics, microcontrollers, and programming, I began studying decision trees, flow charts and graphs. Mostly there were implemented as simple controllers for various autonomous machines, but I was consistently drawn to work with more and more hands-off machines, culminating in the very fun, very satisfying victory in a race between autonomous robots in the computer science department my senior year. By this point, I had yet to branch into any form of automatic learning beyond some weak adaptive control systems. My very first day of graduate school, however, included Machine Learning and from then on, I have been entranced by the process of building brains.

The two most interesting such brains I've made are these:

Murin A machine learning algorithm modeled after operant conditioning which learns to solve procedural tasks with minimal feeback

Dijkstra Planner An automated planning algorithm using Dijkstra's algorithm and a probabilistic state-action model to solve problems

Later on, as I discovered electronics, microcontrollers, and programming, I began studying decision trees, flow charts and graphs. Mostly there were implemented as simple controllers for various autonomous machines, but I was consistently drawn to work with more and more hands-off machines, culminating in the very fun, very satisfying victory in a race between autonomous robots in the computer science department my senior year. By this point, I had yet to branch into any form of automatic learning beyond some weak adaptive control systems. My very first day of graduate school, however, included Machine Learning and from then on, I have been entranced by the process of building brains.

The two most interesting such brains I've made are these:

Murin A machine learning algorithm modeled after operant conditioning which learns to solve procedural tasks with minimal feeback

Dijkstra Planner An automated planning algorithm using Dijkstra's algorithm and a probabilistic state-action model to solve problems