Datathon

Try your hand at big-data analysis!

Are you interested in using big data, analytical techniques, and multidisciplinary teams to uncover new discoveries in public health? The 2021 Datathon is for you!

March 26-31, 2021
co-hosted by the UND Biology club, DaCCoTA, and the Sanford Research's Research Design and Biostatistics Core
One week to analyze a large public dataset related to human health (provided at the beginning of the week)
Cash prizes for 1st, 2nd, and 3rd place Undergraduate submissions.
New to data analysis? Join beginner workshops throughout March!

(Registration is not required, but it does help us to plan for the event and distribute information via the emails that you provide. Feel free to contact Brian Darby at brian.darby *at* und.edu for more information).

Timeline

March 2 & 4, 2021, 6-7 PM (CT) - Beginner workshop on data entry and graphical exploration in R and GGPLOT2 (hosted by Brian Darby).
- [Zoom link will be posted here]
March 9 & 11, 6-7 PM (CT) - Beginner workshop on data entry and neural network classification with Python and Tensorflow (hosted by Kincaid Rowbotham).
- [Zoom link will be posted here]
Friday March 26, 10 AM (CT) - Dataset (with metadata) will be posted on this website and emailed to participants. Brian Darby will host (by Zoom) an (optional) short introduction and Q/A session to introduce the focal dataset, the Datathon rules, and answer any questions that participants might have.
- Click Here for A recording of the Introductory Zoom meeting
Groups have five days to analyze the data and prepare their results for judging.
Results are due by the end of the day Wednesday, March 31.
Thursday, April 1, pre-selected judges will review the submission and select 1st, 2nd, and 3rd place submissions for both categories (Undergraduate, and Other).
Friday April 2, 2021, 3 PM (CT) - Result, awards, and concluding remarks will be presented on a Zoom meeting.
- Click here for the final announcement zoom meeting

Zoom meeting room links will be posted on this website on the week of March 22

Rules and Guidelines

All groups will receive the same large, publicly available, health-related dataset (plus metadata) as a comma-delimited file (expect a 2.5 MB file with approximately 100 variables and 10,000 records).
Groups can be comprised of 1 to 6 participants, and the registration link will allow participants to request organizers to match them to existing groups if so desired.
- Open to any and all who want to participate: Undergraduate, Graduate, Post-Doctoral, Faculty, Professional, etc...
Groups will compete in two categories: Undergraduate and Open.
- Undergraduate groups must have at least 51% of their participants be undergraduate students (within 6 months of graduation and not have started graduate school).
- Open groups have 50% or fewer of their participants as undergraduate students.
- For example, a group of five with three undergraduates would be considered "Undergraduate", while a group of six with three undergraduates would be considered "Open".
Judges will award $500, $250, and $100 to the 1st, 2nd, and 3rd place Undergraduate (only) groups.
- Judges will also rank "Open" group submissions, but without awards.
- Additionally, we will offer $10 pizza gift certificates for the first 25 "preprofessional" (undergraduate, graduate, and post-doctoral) participants that register and participate.
Final submissions will be in the form of a single PDF document in the format of a short (3-4 page) "Note", or "Short Communication".
- There is no restriction on the number of tables or figures, or on formatting of references, but submissions are encouraged to be clear and concise.
Judges will evaluate submissions against a common rubric with five key assessment criteria:
1. Scientific Impact
2. Originality of Approach
3. Appropriateness of Statistical Analysis
4. Clarity of Writing
5. Presentation of Results

Registration

Here is a link to a Dropbox folder that has the National Health and Nutrition Examination Survey (“NHANES”) data that we will be using for this year’s contest:

https://www.dropbox.com/sh/vybtm5q60em5uhz/AACgGm3rX2X49kltSUfYd8kXa?dl=0

Here is a link to more information (and the original source) for the NHANES data:

https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2017