ml-coursera-python-assignments VS coursera-machine-learning-solutions-python

check this out in particular. It's the week 4 homework from Ng's course, redone by someone to be in Python instead of Octave. It's got a built in grader, so you can grab the jupyter notebook, run it locally and it'll tell you when you've got the answer right. I'd recommend taking a crack at it, then when you figure out how to code it, take a look at that micrograd library and see how you could achieve something similar using an object oriented approach.

Python version of assignments which you can submit: https://github.com/dibgerge/ml-coursera-python-assignments

Basically title. I'm halfway thru the original Stanford University Machine Learning course by Andrew Ng, but instead of using the Octave/Matlab exercises, I went with a Python repo. Now, I know the new specialisation course came out and is updated with newer content, more relevant to the state of the industry today. I have the following choices:

A language is only worth learning if it is useful to know. But the only reason 99% of people would learn Octave is just to take that course lol. Besides, (a) the original course can be completed in Python using this repo, and now his new course is actually offered in Python.

FWIW there is a repo you can use to complete the first one in Python. I used it and can vouch that it works perfectly as advertised.

You can do them in python and submit them! https://github.com/dibgerge/ml-coursera-python-assignments

Andrew Ng ML-Coursera Assignments in Python

Last piece... Octave is super easy to get into. I don't personally think it's worth doing Python versions of the homework, but if you really can't stand screwing around with a new language, this repo has alternate versions of the homework to follow that will use Python instead. You can do either these or the original versions, so don't let the Octave scare you. You don't have to use it if you really don't want to, but... like I said, it's not a big deal either way, I just did it in Octave.

Specifically, I'm referring to this github repository: https://github.com/dibgerge/ml-coursera-python-assignments/blob/master/Exercise1/exercise1.ipynb. I'm currently doing Assignment 1.

However, the theta estimates I get using gradient descent appear to be right, as they match the values provided by this person's solutions: https://github.com/suraggupta/coursera-machine-learning-solutions-python/blob/master/Exercise1/exercise1.ipynb.