For collecting references people find useful.

Courses

This course: http://cs231n.stanford.edu/ is fantastic. It covers convolutional neural networks, state of the art research and models (as of 2015 when it was given). There are a lot of ideas and techniques covered in this course that are well worth trying on data at SLAC. Every lecture (14 in all, I believe) is video-ed and available on youtube, for instance youtube: cs231n lecture 9: understanding and visualizing convolutional neural networks

Books/Papers/Websites, etc

...

• https://www.tensorflow.org/
  conda channel for it: https://anaconda.org/jjhelmus/tensorflowtensor-flow is conda-forge
• note, conda channel doesn't have GPU version, get gpu pre-compiled https://www.tensorflow.org/ and install with pip
• Keras: http://keras.io/
• Theano:  http://deeplearning.net/software/theano/ 

...