Dr. Anamaria Berea CDS 490 – Research: Applied Data Science 2020 Summer Semester (3 credits) Mondays (virtually) 12:45 pm - 4:15 pm
Jun 1, 2020, 12:45 PM - Aug 22, 2020, 4:15 PM
This is a research-oriented data science class addressed to seniors and early graduate students interested in exploring unusual datasets, containing images, sounds, natural language, synthetic (computer generated) text or symbols.
Students will embark on data “expeditions” or “adventures”, in which they will try to find hidden clues in the data (e.g., detectpatterns in languages, find repeated signals in audio waves, analyze extinct writing systems, etc.) or use data to search-and-find previously undetected phenomena (e.g., exoplanets in high-orbital satellite imagery, migrations in Earth observation data, detect whether text was computer or human generated through bias and anomaly detection, etc.) or they will have the opportunity to formulate their own EDA and hypotheses based on NASA data repositories (earthobservatory, earthdata, TESS/Kepler, giovanni and ADS). All students will develop their own individual research subproject throughout the semester, which they will be able to reference in their future job applications.
The tools for this class are Tensorflow with Keras (in Colab) and integration of R and Python modules/packages in the same repo/workflow. Optionally, students will be introduced to game theory and to a novel (surprise!) programming language, that leverages both R and Python to higher performances, specific to large data architectures.
Class management follows the rapid research development model from NASA Frontier Development Lab (https://frontierdevelopmentlab.org), but adapted to undergraduate learning, where students will learn how to work in quick turnaround work environments, how to work both independently and in teams, how to communicate research results to a commercial audience and how to develop their own subproject and then integrate with the larger team project. The students will also have the opportunity to meet and interact with researchers from NASA FDL.
*This class will be offered again in the Fall semester with an increased enrollment and a new title “Adventures in Data Science.”