Greg Whittier has been a data scientist at The Aerospace Corporation working in the Machine Intelligence and Exploitation Department since 2019. Greg received a B.S. in chemistry from Georgia Tech in 1993 and a Ph.D. in chemistry from the University of Chicago in 1998. His Ph.D. research was in theoretical chemical dynamics, which mainly involved developing numerical tools and algorithms – what he called “wavefunction engineering”. Between 1998 and 2019, Greg spent 17 years at the Institute for Defense Analyses as a research staff member working in test and evaluation of weapon systems, mainly fighters and UAVs, and three years at the US Army Night Vision Lab working in remote sensing where he had some exposure to machine learning. Prior to OMSCS, Greg had not taken a formal course since graduate school and the last computer science course he took was AP computer science in 1989 when it was taught in Pascal.
"I thought the program could provide some structure to my learning hobby and motivate me to learn more than I would on my own. As a bonus, it might give me a credential that I wouldn’t get with my self-taught route. It was something I could actually do."
As for research, Greg has been working on applying machine learning algorithms to extracting information from remote sensing data. Greg says it is an especially interesting problem to him because it involves modeling both the physics and the data. There is also a software engineering challenge to turning your models into a computationally efficient processing pipeline; thus, it combines a lot of his interests. Greg explained that OMSCS gave him a machine learning credential that he believes made it easier to justify hiring him in his current role despite being mid-career, and he is certain it helped him hit the ground running when he started. Beyond the obviously applicable courses like Machine Learning and Data Visualization and Analytics, Greg has made use of knowledge he gained from the systems courses he took – Operating Systems and High-Performance Computing – when writing data processing pipelines. Suffice it to say, we are incredibly proud of all Greg has and will achieve!
Greg explains the inspiration behind enrolling in OMSCS: "In December 2015, I’d just completed Robert Sedgewick’s Algorithms MOOC and a machine learning MOOC from the University of Washington when I read an article about the first OMSCS graduates and it seemed too good to be true. I thought the program could provide some structure to my learning hobby and motivate me to learn more than I would on my own. As a bonus, it might give me a credential that I wouldn’t get with my self-taught route. It was something I could actually do. OMSCS was affordable and I could do it while working full time with a travel schedule that never would have let me commit to a regular class time."
"OMSCS was affordable and I could do it while working full time with a travel schedule that never would have let me commit to a regular class time."
When reminiscing about his time in OMSCS, Greg acknowledged that what he missed that most was the camaraderie and, as he stated, the “ridiculousness of being on slack at 3am with a bunch of people – who have full time jobs, by the way – trying to get a project done before the midnight Anywhere On Earth deadline”. Though Greg admits it was challenging, he loved the thrill of figuring out the proof in Computability, Complexity, and Algorithms or “finally getting your lunar lander to land without crashing in Reinforcement Learning”. That is what he misses the most about his time at OSMCS: “interacting with my classmates online”. This leads him to advise current students to “get involved on Slack and the forums” and to also not let the fear of difficult courses get in the way of you taking them as what you get out of them is “correlated” with their degree of difficulty.