[Abecadlo]

Sup guys. Im currently doing my undergrad in astronomy, but I know that I willa get no money from it, so I wanted to start learning something profitable. Im pretty good in python so I thought machine learning would be nice. Are there any tutorials or modules you guys would recommend? Ive already searched google but there are so many that Id like second opinion from some1 who knows this stuff

[cxzerios]

Hi,

Two months ago, I did my Machine Learning Course provided by Stanford. I think you should attend that course as the course is a free course. But it doesn't came with Python programming language.

The course explains general explanation of AI, Neural Networks, Machine Learning, and many more, without python examples but if you have the knowledge to use Python, it will helps you a lot as the algorithm explained on the course is very detailed.

Just type on Google this keyword: Coursera Machine Learning by Stanford University

[deadsamdead123]

I agree. Andrew Ng's courses are a great starting point for getting an understanding of basic concepts.

At some point you will want to learn and work in python, but once you grasp enough of the foundational ideas, you can look up tutorials or examples on google for specific types of problems/ use cases.

Libraries like scikit-learn and tensorflow/keras make building ML models almost trivial to the point where anyone with some coding knowledge can just start training and implementing.
To be a good data scientist/ml engineer you will want a fundamental understanding though.

[Junkmannn]

I took a course on ML recommended from Coursera, it was a University of Michigan course "Deep Neural Networks with PyTorch." It's free if you don't need the certification, you just need to click the small "audit this course" link at the bottom when it asks you to pay and you get the entire course for free minus quiz/test grading.

Overall it gave pretty good explanations and being able to go back/forward in the videos on some of the more complicated stuff like CNNs. Couldn't complain about it for the price tag.

I will warn you though, firstly you'll need to do a LOT of independent project work and play with ML if you want to get anywhere near a hire-able level. And secondly Machine Learning isn't for everyone. I learned I wasn't too keen on it myself after devoting a few months to it and although I'm glad I know how it works, that knowledge is pretty unique to ML and doesn't benefit you as a coder in general. My issue is that between the levels of novice and "I worked on AlphaGo" there exists this massive area in the middle where working with ML is more of a trial-and-error "art form" rather than a science, as the inner workings of Neural Nets is WAY too complicated to visualize in your head and you just have to get a feel for it. That's just my experience though.

Good luck!

[eru29643]

As an easier alternative I would suggest to look into django + python. Web development is easier than machine learning and can bring decent paycheck

[jojoba2323]

If you do Andrew Ng's first ML course on Coursera I would also recommend the Deep Learning Specialization. Really good info there

[flipment47]

You should learn how to code before you attempt to get into Machine Learning type stuff. Otherwise you'll just be yelling at the interpreter. You should also know that depending on what you're actually doing, you might need to buy expensive hardware to accelerate your Machine Learning.