Materials for TensorFlow is Hard, Keras is Not talk during the TensorFlow Summit Extended KL event at iTrain on 18th February 2017.

The talk was to introduce Keras - an easy-to-use high level neural network API library, for you to use on top of TensorFlow.

Click to read:

• Food Image Classification Demo Notebook
• Slides

If you have done codelab and found TensorFlow-MNIST is hard to follow, try to read and understand this easier one with Keras:

• Keras Examples - MNIST CNN

If you are new to machine learning, these are the following machine learning courses you can take or read:

• CS231n explanation of convolutional neural network
• [SERIOUS ONLY] Andrew Ng from Stanford's machine learning Coursera

Got questions or issues? Open a GitHub issue in this repo, I will try my best to answer!