How much Python is required for the fast.ai machine learning course?

I would like to take fast.ai’s free online course Practical Deep Learning for Coders. You need to know Python to do the course; that’s why I started Learn Python The Hard Way. Has anyone here taken the fast.ai course?

I’m wondering if Learn Python The Hard Way will be sufficient. Will I need to also do Learn More Python The Hard Way or another course after this? I’m just curious what the path ahead of me looks like. Thanks in advance to anyone who can help answer this question.

I’ve not taken these courses but actually reading the info on the site gives some indication:

It suggests a years coding experience so I’d assume you need more than just LPTHW, and probably some exposure to algorithms and data manipulation. I expect that covering LMPTHW (More Python) would be a good investment, but ask yourself this…

Do I want to learn just enough so I am trying to learn a new topic AND the Python language in parallel, or would I prefer to be competent in Python so I can learn implementation of new libraries/concepts with focus?

I know which I would choose, as I did exactly that before moving into data science

1 Like

What did you do? Do you use Python for data science? How did you learn it?

I use Datacamp as once completing LPTHW the basic concepts and libraries for extraction, analysis and presentation were quite straight forward.

There is a lot of content but this is junior level stuff compared to the full path ahead. But you gotta start somewhere.

Can you explain to me the path you took? First, you took LPTHW to learn Python. Then, you took DataCamp’s Data Scientist with Python course. Is that correct?

What worked for you might not work for me, but I’m curious to know how others have done it.

To be honest my path was much longer than that but in terms of Python and data science, yes I studied LPTHW, then studied some other Python books on web development, design patterns, testing and eventually realised data science was a good fit.

I signed up with DataCamp as it was quite cheap and started following one of their ‘learning paths’ for Data Scientist in Python. ( The paths are essentially an ordering of modules that make sense in a given context).

But it’s not a case of 1. Learn Python. 2. Learn Data Science. Both are vast subjects depending on where you go with them. I’m actually attending a Meetup locally with some Scientists to get some steer on what ‘they’ think is essential knowledge rather than what the course venders want to sell.

But a quick dabble into some more professional analysis shows good knowledge of statistics and sociology is as important as knowing how to code.

1 Like

Sociology?! Really? :open_mouth:

Of course. Why else are we machine learning? Actually human cognitive bias is the reason AI will never be given the responsibility levels above a child. Although it’s being implemented in areas such as insurance evaluation and even criminal sentencing, the social biases already exist in the dataset and render AI a very clever ‘child’ in terms of responsibility. It only knows what it’s been told, and that data is biased already.

2 Likes