After finishing the Coursera ml-course I want to apply the learned techniques using Python. To get started, I'll make a 1-to-1 translation of the examples using Numpy.
All data files from ml-course are missing; I'm quite sure that making this code translation available publicly doesn't violate the Coursera honor code (or does it? please drop me a line if it does!) but I'm somewhat unsure about making the data files available too.
I'd be very thankful for feedback from more experienced Numpy users about the very many ways I'm abusing Numpy or any tips on how to use it more properly. Same goes for the Python code. Please note that the code duplication is intended; the original Octave code contains a lot of duplication and I want to preserve the general lecture structure.