Deep Water

Loading...

From the course by H2O

Practical Machine Learning on H2O

17 ratings

Explore Related Videos

At Coursera, you will find the best lectures in the world. Here are some of our personalized recommendations for you

H2O

Practical Machine Learning on H2O

17 ratings

In this course, we will learn all the core techniques needed to make effective use of H2O. Even if you have no prior experience of machine learning, even if your math is weak, by the end of this course you will be able to make machine learning models using a variety of algorithms. We will be using linear models, random forest, GBMs and of course deep learning, as well as some unsupervised learning algorithms. You will also be able to evaluate your models and choose the best model to suit not just your data but the other business restraints you may be under.

From the lesson

Everything Else!

Pojo And Mojo5:49

Driverless AI1:58

Meet the Instructors

Darren Cook

Deep Water, so Deep Water is another H2O technology.

The main H2O software we've been looking at in this course has a lot of power.

But as I'm sure you're aware, there are alternative competing machine

learning technologies such as TensorFlow, MXNet, Caffe, and so on.

Each of those has its pros and cons.

The idea behind Deep Water is to wrap them,

So as to allow you to use TensorFlow, MXNet, Caffe, but

still stick within the rest of your H2O infrastructure.

So you get to use them as a back end.

Now there are a few challenges with this.

For the H2O developers, it's quite the challenge to

be mapping a back end as a moving target because those libraries keep updating.

They keep releasing new versions.

This is also related to another downside, which is that Deep Water is

quite difficult to install if you do it yourself.

You need to be able to install each of those components, TensorFlow,

MXNet, Caffe, get them working with GPUs all on the same machine,

as well as installing Deep Water.

The good news is H2O provides some ready-made setups.

So there's an Amazon AMI machine image.

There's also a Docker image that you can use.

So those solutions help you bypass a lot of the setup complexity.

There are two main reasons you might want to use Deep Water.

The first is so you can make use of GPUs in your deep learning in particular.

The second is to use ready-made models.

So this comes back to another pro and con.

H2O, in my opinion,

makes creating deep learning models just about as easy as it possibly can be.

The same can be said for the alternative technologies.

They tend to expose a lot more of the inner details.

You need a lot more knowledge about how the algorithms

actually work to be able to use them.

But if someone has already made a very useful model, in TensorFlow for instance,

say a very deep convolutional network for image recognition that you want to use.

Deep Water allows you to bring in that model and use it, but

still stick within your familiar infrastructure,

whether it's Flow or the H2O client for R and Python.

So if you want to learn more about H2O Deep Water,

there are some links on the documentation page docs.h2o.ai.

That's the best place to start.

Explore our Catalog

Join for free and get personalized recommendations, updates and offers.

Coursera provides universal access to the world’s best education, partnering with top universities and organizations to offer courses online.

© 2019 Coursera Inc. All rights reserved.

Download on the App Store

Get it on Google Play