This isn’t a typical post. Though I prefer working in R, I also do quite a bit of work in Python these days. I really like the way Python is setup for machine learning and there are some deep learning focused packages that just don’t exist on R yet. One of these is a package called Fastai. Fastai is a python package that wraps the Pytorch deep learning library from Facebook. I learned about it while taking the Fastai MOOC’s on machine learning and deep learning and I’ve wanted to play around with it myself ever since.
To give it a whirl, I spent some time building a Kaggle Kernal that uses the Fastai package(along with sklearn) to build a random forest to predict interview attendence. While I know there are ways to integrate both Python and R posts into a blog, since I currently blog using blogdown from R, I’m still figuring out the details behind posting a Python notebook here. For now though, I’ll link to my Kaggle Kernal and apologize for the janky work-around.
Anyway, you can find the kernal here. Enjoy!