Summary of PhD Research and Time Series Modeling Experience:
During my PhD, I used spectroscopic data from ground-based and space-based telescopes to determine the atmospheric properties of planets light years away from us (i.e., exoplanets). I did this by spectroscopically observing the host star before, during, and after the planet passed in front of it (i.e., transit). This allowed us to compare the spectroscopic profile of the stellar light attenuated by the planet's atmosphere during transit and the stellar light alone (i.e. transmission spectroscopy). With this, we deduce what atmospheric properties the planet has. Given that the transit lasted hours and I had to properly model the full transit features, I developed strong skills in analyzing time series data.
Experience with Remote Sensing Data, Machine Learning, and Bayesian Statistical Tools:
When using ground-based telescopes, the observations were contaminated by signals from our atmosphere, the planet's host star, and instrumental and optical systematics. Thus, I used machine learning algorithms such as Gaussian processes, principal component analysis, and polynomial regressions in conjunction with probability distribution samplers such as Markov chain Monte Carlo (MCMC) and nested sampling to model out systematics. Through this, I have a strong grasp of the machine learning and statistical procedures needed to thoroughly reduce and analyze remote sensing data.
Experience in Quantifying Prediction Uncertainties:
In the academic community, it is vital to clearly outline potential biases and reasonably estimate uncertainties. Overestimating uncertainties could squash significant signals, which is especially the case in the low signal-to-noise regime that most of my ground-based data were in. Conversely, underestimating uncertainties can easily lead to inaccurate interpretation of features in the data. As such, I invested a lot of time working on how to properly measure uncertainties in my machine learning (ML) algorithms. I published a peer-reviewed research paper (McGruder, et. al. 2022. See publications) that discusses the benefits of different detrending algorithms and performs statistical tests of the biases and estimated uncertainties of the different algorithms using data I synthesized. In this work, I also created a new detrending algorithm for my data (see github), which was applied in two subsequent research papers (McGruder, et. al. Sept 2023 and McGruder, et. al. Nov 2023. See publications).
Leadership Roles and Public Speaking:
I obtained project management skills as the lead of 6 ground-based telescope proposals, 5 published peer-reviewed papers, and 4 space telescope proposals. In each of these projects, I collaborated with and led teams of scientists from around the world, in different career stages, and from varying backgrounds. I also have demonstrated strong public speaking skills with 21 public presentations and/or talks given throughout my career all over the world (see presentations).
Hands-on Observational Experience:
I would often travel to Chile to conduct ground-based astronomy observations at the Las Campanas observatory in the Atacama Desert. My responsibilities included the instrumental setup, calibrations, and mid-observation troubleshooting.
