Hacker News new | past | comments | ask | show | jobs | submit login
Kaggle Learn review: there is a deep learning track and it is worth your time (techandmortals.wordpress.com)
219 points by akashtndn on Jan 27, 2018 | hide | past | favorite | 12 comments



Since the "How Do I Learn AI/ML" question pops up on Hacker News once a month (most recent: https://news.ycombinator.com/item?id=16138353), here's my comments:

Yes, Al/ML MOOCs teach the corresponding tools well, and the creation of new tools like Keras make the field much more accessable. The obsolete gatekeeping by the AI/ML elites who say "you can't use AI/ML unless you have a PhD/5 years research experience" is one of the things I really hate about the industry.

However, contrary to the thought pieces that tend to pop up, taking and passing a MOOC doesn't mean you'll be an expert in the field (and this applies for most MOOCs, honestly). They're very good for learning an overview of the technology, but nothing beats appling the tools on a real-world, noisy dataset, and solving the inevitable little problems that crop up during the process.

Reviewing the Keras documentation (https://keras.io) and examples (https://github.com/keras-team/keras/tree/master/examples) are honestly much better teachers of AI/ML than any MOOC, in my opinion.


I'd suggest fast.ai mooc before keras docs. I took Hintons course and tried to learn through tf/keras docs, but wasn't able to really get going until I found fast.ai. Some of the best "classes" I've ever watched and there's a ton of people helping in the forums.


While I agree that PhD gate-keeping is frustrating, I've found a sizable subset of the people who say that really mean "we want you to have the mathematical foundation for this", not "we require a PhD". I don't have a PhD, but I've found that generally, as long as I can show I have the theory down, employers don't seem to mind.


The PhD reference in the comment is more toward other comments made on Hacker News/Reddit/Medium thought pieces.

For job hunting, the credibility issue is more nuanced, and I wrote a separate rant about that a couple weeks ago: https://twitter.com/minimaxir/status/951117788835278848


I do agree that for applying technologies in real world problems, a much deeper understanding is required than what majority of MOOCs provide. That being said, different learning methods suit different type of learners. For some, starting out with a hands-on overview of the topic at hand works best. This is where tutorials such as this one shine through.


> Reviewing the Keras documentation (https://keras.io) and examples (https://github.com/keras-team/keras/tree/master/examples) are honestly much better teachers of AI/ML than any MOOC, in my opinion.

Wholeheartedly agree. It’s simple to setup a Bidirectional LSTM and run it with PyTorch if you are trying to predict time series data.


I'm the lead on the Kaggle Learn project, and the author of the deep learning track.

I'm happy to answer questions here.

I agree with the commenter saying you need to do your own projects to understand these topics.

Our deep learning track is meant to be the fastest path to knowing enough to do your own projects. You can do the entire track, including the hands-on exercises, in a single sitting.

We won't make you an expert in an afternoon, but you'll know enough to start doing your own projects. For most people that's also the point where Deep Learning becomes fun enough that you'll find time to keep learning.

Kaggle Learn is still in a very early stage. We'll add more lessons soon. But we'll stay committed to the goal of getting you up-to-speed quickly, so you can take on your own projects.


A very nice playbook style breakdown of steps to ease in to Machine Learning and eventually Deeplearning Of should I call it Differential Programming??? Keep it up and hope to see more of your writing. May I also add the very approachable book by Tariq Rashid https://github.com/makeyourownneuralnetwork.

Ananth


Owing to my inexperience with the topic, I am not sure if the term 'differential programming' belongs here. However, this book looks promising. Thanks.


@akash - good article. Love to see some more in-depth stuff next time !


Thanks and definitely.


Thank you ! I was thinking to put time into deep learning and it seems like this is a good way to start !




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: