This past summer I had the wonderful opportunity to work with The Yaron research group based out of Carnegie Mellon University. The project was focused on using machine learning to create low-level models that replicate high-level data corresponding to the electronic structure of functional groups. The first half of the internship focused on learning the programming language Python and the Hartree-Fock method which approximates a solution to the time-independent Schrödinger equation. The second half of the internship was spent studying the effects of swapping out certain elements in the high-level data with low-level data to determine what portions of the model would be suitable for machine learning. I hadn't taken physical chemistry yet, but fortunately I had taken Professor Sanov's chem 380 course which proved to help tremendously with the mathematical aspect of the project. Since the project was computational in nature it allowed me to work from anywhere. Group meetings and daily communication was done through Skype while results and data was shared through GitHub.
With the flexibility of the internship I traveled to Sedona and Flagstaff for rock climbing. I camped with a group of friends and periodically went into town for Internet to check in with the research group and upload data results. This may be the most unconventional internship heard of, but I learned a lot while still being able to enjoy the summer. I'd like to thank Professor Yaron and his graduate students Christopher Collins and Matteus Tanha for their willingness to teach, insight, and patience.