Outcomes

D3EMs learning outcomes with brief reflections:

• Interdisciplinary Knowledge Generation
  • Undoubtedly, D3EM has contributed to my ability to generate thoughts in other fields. Knowledge of materials science has created extremely unique application domains in which I have been able to conceive ideas I never would have been able to conceive before. I have the materials science expertise, and the expertise of those researchers in the department, to thank for the opportunity and collaborations
  • One of my biggest moments recently came from a meeting with a materials scientist who was describing a dataset they had just collected to me. It had all the components I needed to train an algorithm I had been conceiving for a while, and I didn't even know the data to validate it existed!
• Collaboration
  • I regularly collaborate with the teams I have formed through the years in D3EM. At least monthly, in some cases weekly, we get together to bounce ideas and discuss results or ways forward. Some meetings are better than others, naturally, but the collaborative nature is there, and
• Conflict Resolution
  • For the most part, in my time in D3EM, I haven't need to really dip into conflict resolution on an interdisciplinary team. By-and-large, our team of interdisciplinary collaborators gets along well. We do have differing backgrounds, and we all do grow through conversations. Sometimes a misunderstanding can be cleared up through constructive discussion, making a concept more clear or paving a new path forward.
• Oral Communication
  • In our D3EM learning community we were asked to explain our research in 2 minutes, to a group of people that don't understand it. This was a fantastic exercise in oral communication. I talk a lot (so writing blog posts is perfect for me) and I have a tendency to over-share of over-explain. Therefore summarizing research in 2 minutes or less is a fantastic opportunity to fine-tune those skills.
• Written Communication
  • Its extremely interesting to go back and view some of my writing before I was in D3EM. That was 3 years ago, as I was beginning my Ph.D., so one could surmise that the growth in my writing was a natural process. However, I think that taking part in the writing community early on was helpful in framing and providing a good foundation for my writing. More specifics can be found in my writing community reflection.
• Self-Reflection
  • Self reflection was never big for me. However, as I've progressed in my degree I've grown more and more fond of writing, and sharing those opinions. Interestingly enough, the DL community is vibrant with blogs and has its now active medium channel, for which users can contribute. A lot of this writing on Towards Data Science (a medium subchannel) is technical - i.e. "how to build a VGG16 CNN in Tensorflow" - but quite a bit is introspective - i.e. "10 things I wish I knew as a data scientist." Prior to my time in D3EM, I probably never would have volunteered my time to contribute to that community, and now I have a running list of topics I'd like to write about.
• Ethics
  • Ethical research is a must. In the DL community this can prove some level of difficulty. We have hyperparameters for all our models, parameters that, if changed, generate a new model, with different depth or width, training algorithms with different learning rates or time steppers. All of these can be modified, and yield a different results. However, these new results are not publication worthy. This is hard for some, but the real authority in DL research come not from a hyperparameter space search, but comes in the form of a new architecture that solves a problem. Or a new way of training an existing architecture. Or a new normalization layer that results in better stability in training. These are hard problems. Ethically, these are the problems that we should be tackling - not publishing and "epsilon" paper, in which a single or set of parameters are changed. It is up to us - the researchers - to determine whether a contribution is really a contribution.
• Interdisciplinary Research
  • Interdisciplinary research has major benefits, albeit some minor drawbacks, that I have had the opportunity to experience over the last few years. Reflections on research are outline in the research experiences reflection section.
• Multidisciplinary skills
  • This one is big for me, as I've come a long way in things that I never would have touched with a traditional DL background, such as phase field modeling (encompassing thermodynamics and kinetics) and numerical methods. These topics are prominent in a lot of engineering fields, such as aerospace and mechanical engineering, but really electrical undergrad (at least at Texas A&M) doesn't include these topics in its curriculum. This has been an interesting experience, getting to learn all these topics that are very fundamental engineering topics that my discipline really doesn't get too much experience with. Although, it does take time to study and understand these topics at a research level, a topic discussed in my research reflections.
• Materials Science Engineering
  • Materials science and engineering has truly broadened my understanding of engineering as a whole. As discussed briefly, I've been able to learn a lot about other engineering disciplines that I never would have touched with a typical electrical engineering background. Not only was i able to study these, but being able to contribute to the body of knowledge in these fields has required a level of expertise that I'm still seeking. They say learning is never done, but to be an engineer that wears this many hats is to always be reading, writing, or coding in some area of science.
• Informatics
• Design