Closing Out 2017
I took advantage of some open space in my work schedule to really crank through the online class on Udemy, learning the basics of Pytorch for deep learning. It was a well constructed class, and balanced the theory and mathematics behind the concepts with the practical, coding part of building neural networks in PyTorch. I definitely feel comfortable using this class as a good jumping off point for creating my own neural network models.
Now that I have completed this class, I want to turn my focus toward putting these skills to work on Kaggle competitions, or projects of my own choosing.
New Project Idea
Speaking of new projects, I was thinking about how to practice with PyTorch and neural nets, and decided that a satellite imagery project would be a good place to start. While mulling over options for this project, I thought about how I could potentially connect it to work that my brother is doing in his city - Flint, Michigan. As it turns out, he was already thinking about adding a data component to his ongoing work in Flint, as they redevelop and recover from the economic downturn of the mid-00s. As of this writing, my plan is to research semantic segmentation of satellite imagery, applying that to Flint, perhaps looking for cities around the United States that are similar and presenting that as information that could help generate ideas for how to redevelop currently unused and underused areas of the city. Early days for this project, as they say, but I hope to get moving on this quickly in the new year.
Not too much to report on the job front. A coworker who has been in the same/similar boat as me ended up getting a new job before me, albeit at a different company and requiring a cross-country move. To be frank, it’ll be addition by subtraction for me, as this person was contributing to the toxic environment at work. Other than that, I have an application in with another department at work that’s in the facility nearer to my house. It’s a stepping-stone job (as they all seem to be), but it’s something. I hope the change of scenery and work will re-energize me. No idea if I have an interview, but things tend to slow down with people taking vacation as the year comes to an end. I imagine I’ll find out one way or the other soon after we return in January. In the meantime, I am actively working on applying the data science and machine learning skills to my current role, as I bolster my resume with projects in data science. This journey is taking longer than I was hoping, but at least it’s moving on a sure trajectory at the moment.Share on Twitter Share on Facebook