Wrapping Up Udacity's Deep Learning Nanodegree

Written on July 22, 2017
[ udacity  misc  wwe  ]

Udacity’s Deep Learning nanodegree was such a great experience, and I learned so much about Python, TensorFlow, Keras, and – of course – neural networks.

Before I enrolled, I remember being asked by someone genuinely interested in the topic, “Do you know anything about neural networks?” I said, “Sort of,” to which he replied, “I literally have no idea what they are or how they work.”

That got me thinking: Do I really know what neural networks are? I told him, “Yea, I get the gist, but I guess I don’t really know much about them.”

Boy, was I right!

The next morning I had an email from Udacity promoting the first run of their new nanodegree foundation on deep learning. The timing was perfect: I enrolled immediately.

My background in physics surely gave me a mathematical edge and insight. It was fun to think of the unfamiliar constructs and terminology in terms of things I knew from my past: topology and manifolds, Lagrangians and Hamiltonians, time series and spectral analysis!

Though I had dabbled in Python previously (and it being the language I originally began learning to code in way back!), I was really an outsider to it before this class. Most of my data analysis and modeling research in the late 00s was in MatLab, and sometimes IDL. When Coursera first came out, my brother and I enrolled in its original courses on R by Roger Peng and Jeff Leek, and I spent quite a few years cultivating my R chops and trying to convert anyone who would listen. However, most of my colleagues used IDL and, having productively used it for 10+ years successfully, were not interested in any newfangled languages.

Though I may have been an outsider before the nanodegree, I am now nearly 100\% pythonista. Before, I was doing all my modeling, statistical analyses, reports, and presentations in R/RStudio. After, R has become somewhat of a “library” that I sometimes call from within a Jupyter Notebook (however, the more comfortable I become with the equivalent python libraries, the less this happens).

Anyway, point is: I can’t recommend this degree enough! It was challenging to keep up with at points because of other things going on in my life (e.g., kid, marathon training, work!), but overall it was fun and well worth the effort. Just the insight into modern technolgies from Google, Amazon, and Facebook, is alone worth the cost.