I recently finished MLS Course 1 and eager to start the rest of the courses, yet, I wish to get more coding practice related to the topics.
I took notes and understood the overall theory
I understood most of the code from the practice labs
I somewhat could do the assignments on my own by reviewing my notes
My concern is, how can I improve my coding skills so that I myself can create a practice lab or assignment as a project starting from scratch. I feel stuck whenever I start a blank notebook, I don’t know how to even start, yet when reviewing labs, I understand 80% of the provided code. Any suggestions?
Yes, totally agree with @TMosh. I self-learned almost everything, but before I started to do data science I had already known programming (C, C++ - but I self learned them too) so picking up Python wasn’t that difficult, and the only thing is to get familiar with it, the syntax, the packages, and so on, and in this sense, I was like you that at the beginning, when I opened a jupyter notebook I didn’t know what to do except maybe printing “Hello World!”. If you already have some programming experience, I would really just suggest you to set a goal, for example, you download a dataset and you ask yourself to use xgboost to build a model and make use of all the skills you have learned. Don’t insist to start from scratch, and it’s ok and actually better for you to copy-and-paste some ready code and there can be many different versions of it available on internet, but you need to test the code by changing it, and try to find other ways to rewrite the code, not necessarily optimize it, but just to check what you are able to do.
I think keep practicing like this would be a good start. However, if you don’t have too much experience on programming but you are eager to try, you can actually do the same as what I described above, because your learning pace is set by you and by your persistence. Python is designed to be easy to use, and many millions of people know it, so why not you?
I’ve been wondering in the data science/machine learning space for around 5 months now, but it is overwhelmingly tiresome to see too much information and no clear path as what or where to chose from. Any recommendations as on how or where I could practice or read more about it?
PS. This is one of the first courses after trying around 10-15 in udemy, udacity, edx, that finally caught my attention and feel like I can learn from.
If you spent 5 months and you can’t find a clear path or a course, then please stop looking for it. I have seen many Python courses teaching us how to write a for loop or how to format a string, and if you are interested in those, you can find it on coursera, otherwise, just start getting your hands dirty and set a goal. Here is a place you can find opportunities to practice, not taking courses, but practice:
You see in below there are a few categories, visit “Getting started” or “playground”, think about what you have learnt, find a topic you like, check out others’ code.
Raymond posted the Kaggle recommendation before I could
Another way to get data to practice on your own is an online repository. There are many of them, and they’re free, and have pre-made data sets ready for your experiments.
The “UCI Machine Learning Repository” is a good place to start.
Brother, will you help me in this project" extract trends from social media data " .
please do understand me what to do and how should be my approach in this field hope you will help in this project .
thanking you !
Hello Tushar, thank you for asking. There is a “General Discussion Category” in this community, and i suggest we can discuss your experience of your project there. Maybe you can share what you have tried, your finding, and we can disucss how to make use of your finding for some decisions. You don’t need to share your code and your dataset, but if you think some toy code and toy dataset might be helpful for the discussion, please feel free to share.
If you want to make sure I will be notified of your posts, please tag me (@rmwkwok) there.
