Update Referenced Jupyter Image(s)

[kilbd]

In the included base image Dockerfiles, the referenced scipy-notebook Jupyter image is about 18 months old. When I tried to conda install two packages in my custom Dockerfile, it wanted to downgrade Python to 2.7, and eventually exited with this error:

Traceback (most recent call last):
  File "/opt/conda/bin/conda", line 7, in <module>
    from conda.cli import main
ImportError: No module named conda.cli

When I removed the :135a595d2a93 tag reference to implicitly use :latest, the build went fine. I assume the miniconda version in the old image has a bug. I did not test this with the other 2 Coursera base images.

I understand not wanting to use :latest, and Jupyter doesn't have the friendliest tagging scheme on Docker Hub, but you could at least mention using a different image tag in the README when you encourage using your images as a base for customization. Better still, if you really want folks to use your default images as a base for customization, then publish your images to Docker Hub, perhaps with a nicer tag/versioning scheme.