H2O vs. Fast.ai

My colleague told me about H2O and I am wondering how H2O.ai compare to Fast.ai. Looks like a lot of firms are already using H2O .

  • Is H2O much easier to use (looks like it has graphical interface for learning etc.)?
  • Is H2O skills more helpful in creating useful applications?
  • Should I learn H2O instead of Fast.ai?

Any guidance will be very helpful.

It seems to be comparing apples to oranges. H2O looks from the website to be some sort of suite of tools for the enterprise and the fast.ai library is sort of like sklearn or multiple R packages with the difference being that it is a cutting edge deep learning library with a very appealing interface.

I have no clue what you should do and I don’t think anyone on the forums can tell you that :slight_smile: Not sure where in your machine learning journey you are but the best piece of advice I can give you is that probably the tools don’t matter much - you should focus on what builds up your understanding of the techniques the most. For me the answer is and has been very simple for last couple of months - fast.ai. And being able to use cutting edge techniques is also nice :slight_smile:

All the best in your search wherever it may lead you :slight_smile:

1 Like

I’m under the impression that after you learn the concepts from fast.ai you will be able to pick up whatever DL library you like. Fast.ai is much easier to get started with compared to other DL libraries (e.g keras).