Because correlation only gets you part of the way there

The BD2K Center for Causal Discovery, a collaboration among the University of Pittsburgh (Pitt), Carnegie Mellon University (CMU), Pittsburgh Supercomputing Center, and Yale University, is holding their second annual Datathon designed to instruct and challenge biomedical researchers on the use and application of causal modeling and discovery (CMD) tools in a “bring your own data event”.  

We will invite participants to bring their own data to the event and offer cash prizes for the best analyses and results. As a prerequisite, Datathon participants will be expected to have attended a CCD Short Course or other training session, download CCD software and perform preliminary data formatting to accommodate the time frame available. We will provide the formatting specifications for their data so that they can prepare their data in advance. Participants will be required to use at least one of our CCD tools for their analysis: causal web application, causal command application, Tetrad Desktop, or causal apis (Java, R, Python).

Feel free to look at last year's datathon page to see what participants worked on: https://causal-discovery-datathon-3582.devpost.com/

View full rules

Eligibility

Scientists in the fields of clinical informatics, bioinformatics, and general data science as well as diverse biomedical and clinical research disciplines are invited to our datathon.  As a prerequisite, participants should have taken our Summer Short Course in Causal Discovery or other CCD seminar.  If you haven't taken one of our courses/seminars, you're in luck, the annual short course immediately preceeds the datathon!  See here for information about the short course.  Participants for the datathon should register at the short course website as it is a prerequisite (unless you have attended previously)

Requirements

Participants will prepare a short slide presentation (in person presentation optional) on the results of their analysis so that their entry can be reviewed by our panel of judges.

Judges

No avatar 100

TBA

Judging Criteria

  • Big Data
    Size and complexity of data (10 points)
  • Impact
    Impact with regards to the causal hypotheses generated (10 points)
  • Innovation
    Innovation in the use of CCD tools (5 points)