Last time we wrote an image classifier using TensorFlow for Poets. This time, we’ll write a basic one using TF.Learn. To make it easier for you to try this out, I wrote a Jupyter Notebook for this episode -- -- and I’ll start with a quick screencast of installing TensorFlow using Docker, and serving the notebook. This is a great way to get all the dependencies installed and properly configured. I've linked some additional notebooks below you can try out, too. Next, I’ll start introducing a linear classifier. My goal here is just to get us started. I’d like to spend a lot more time on this next episode, if there’s interest? I have a couple alternate ways of introducing them that I think would be helpful (and I put some exceptional links below for you to check out to learn more, esp. Colah's blog and CS231n - wow!). Finally, I’ll show you how to reproduce those nifty images of weights from TensorFlow.org's Basic MNIST’s tutorial.
Jupyter Notebook:
Docker images:
MNIST tutorial:
Visualizing MNIST: (this blog is outstanding)
More notebooks:
More about linear classifiers:
Much more about linear classifiers: (this course is outstanding, highly recommended)
More TF.Learn examples:
Thanks for watching, and have fun! For updates on new episodes, you can find me on Twitter at www.twitter.com/random_forests