Skip to content

Latest commit

 

History

History
executable file
·
78 lines (51 loc) · 2.47 KB

HOWTO-1.md

File metadata and controls

executable file
·
78 lines (51 loc) · 2.47 KB

HOWTO 1 of the Machine Learning Committee of SIIM.

Intended Audience:

People who are brand new to machine learning, and don't want to acquire a Ph.D. in computer science just to start learning

Table of Contents

  1. Background
  2. Do you have the required hardware?
  3. Getting and installing Docker
  4. Getting and running a prebuilt Docker from Dockerhub

Chapter 1

SiiM established the Machine Learning COmmitee to accomplish several goals:

  1. teach concepts to people new to the area
  2. encourage learning and modifying from existing turnkey appliances
  3. advance the field with teams of people who are willing to co-develop on reproducible platforms

Towards these ends, the Comm. has settled on the idea of encouraging participants to contribute their code in the reproducible format known as Jupyter notebooks

Chapter 2

Since we are doing machine learning, and most ML platforms require a GPU, the first thing to figure out is does your computer have the hardware called out by the Docker author that you intend to use?

For the purpose of this HOWTO, we will assume the user is going to pull down and run the DOcker based on Dr. George Shih's Keras based Docker. The link to that (and its requirements) is here

https://github.com/ImagingInformatics/machine-learning/tree/master/docker-keras-tensorflow-python3-jupyter

From his README file looks like all we need is: 4 GB of RAM, a normal CPU, and a modern OS. Moving along ...

Chapter 3

Again, for the purpose of this HOWTO, we are going to use the Keras on Jupyter Docker that already has an entry on DOckerhub. So all you have to do is have the Docker run-time engine, and a Dockerhub account, to pull down the premade Docker and run it

First, let's install the Docker engine for your PC. You can get the Docker runtime for your computer here https://www.docker.com/

More specifically :

  1. Windows 10 Pro and newer (https://hub.docker.com/?overlay=onboarding)
  2. Windows 10 Home or Older Windows (https://docs.docker.com/toolbox/overview/)
  3. MacOS >= 10.12 (https://hub.docker.com/?overlay=onboarding)
  4. Linux: varies by distribution
  5. iOS or Android: out of scope

Follow the instructions for your OS and install it

Chapter 4

Now that we have the Docker engine installed, we refer you to the instructions at https://github.com/ImagingInformatics/machine-learning/tree/master/docker-keras-tensorflow-python3-jupyter

to fetch and run your first example of a Jupyter notebook tailored for Machine Learning. enjoy !