Applied Machine Learning for Psychiatry/Neuroscience

Two-day machine learning workshop at Duke University Medical School & Durham VA

Adam Chekroud, Yale University

---

• This repository contains the materials that will be used during a two-day machine learning workshop at Duke. This workshop will serve as a theoretical and practical introduction to using machine learning methods to solve clinically important problems. It takes a high-level approach, with almost no equations, and uses examples drawn from various areas of biology.

• The materials are all available for free (yay, science!), but please do cite our paper and let me know if you find any of this useful!

• You will find folders inside containing:

  • an example data set
  • talk slides
  • PDFs for relevant textbooks
  • other resources.

There are also some scripts

1. basic_R.R

• Contains some very introductory R code.
• If you have used R before then you won't need to read it
• If you are new to R then it may help give you an idea of syntax

2. dukePackageInstaller.R

• Should install the main R packages we will be using.
• You should be able to source the script directly, and it will indicate whether it was likely to work
  • to do this type source dukePackageInstaller.R into your terminal.
  • if it worked, it should return "This probably worked as expected"
• Either way, if you read the code it should be fairly easy to see which packages you need to install manually

3. skeleton.R

• Contains the main demonstration script that we will be working through.

As a reminder:

• You can install the latest version of R for free here
• And install the latest version of Rstudio for free here

Please feel free to email me if you have any questions, or if you spot anything that isn't working: adam dot chekroud at yale dot edu

The appropriate citation for these materials is:

Chekroud AM, et al. (2016) Cross-trial prediction of treatment outcome in depression: a machine learning approach. The Lancet Psychiatry 0366(15):1–8. Available here

Please do not share or reuse these materials without permission. I have no copyright, nor do I have money to pay lawyers to do anything about it, but I will be sad, and tweet everyone I know to tell them you were naughty.