.png&w=3840&q=75)
.png&w=3840&q=75)
Datathon · 3rd Annual
MD+ 2024 Datathon Multiple competition tracks.
The 3rd annual MD+ Datathon — a national, month-long event, and the first to run multiple parallel competition tracks each using a different publicly available dataset.
- Dates
- 10/23/24 – 11/13/24
- Pitch competition
- Mon 11/18/24 · 6pm EST
- Location
- Remote
- Prize
- $3,000 grand prize
About
The 3rd Annual Datathon.
Welcome to the 2024 MDplus Datathon! The 3rd annual MD+ Datathon was a national, month-long event hosted by MD+ and sponsors to foster innovative thinking about complex healthcare problems and their data-driven solutions.
Participants worked with medical students, graduate students, and trainees from all levels to generate insights and engineer solutions from clinical datasets. In contrast to prior years, this datathon was divided into three separate competition tracks, each using a different publicly available dataset. Specific datasets for each track were announced after team formation and made available on Hugging Face.
Final projects and presentations were reviewed by a panel of expert judges, and the top 8 projects were invited to a live pitch competition. Pitches were capped at 8 minutes followed by 2 minutes of Q&A.
Competition tracks
Mental Health Track
Clinical Documentation Track
Medical Education Track
Use of generative AI
If you used an LLM as part of your project, you could only use open-source models with 8 billion parameters or less. This restriction applied to the final submitted product, not the tools used during development.
Timeline
Key dates.
- Tue 10/8/24
Signups open
- Fri 10/18/24 · 11:59pm AOE
Signups close
- 10/18/24 – 10/23/24
Application review, notification, and team formation
- Wed 10/23/24 · 6pm EST
Start of datathon, datasets released
- Wed 11/13/24 · 11:59pm AOE
Final submissions due
- Mon 11/18/24 · 6pm EST
Finalist presentations in live pitch competition
FAQs
Common questions.
I have little/no data science or computational experience. Can I still participate?
Absolutely! Learning the computational tools is half the fun. Participants have access to tutorials covering the basics of Python and R and how to analyze the dataset. Judges care more about the insights derived and the data analysis than computational novelty or complexity.
What will we actually be doing?
Students are provided a dataset (e.g., claims and hospital data) and asked to identify an addressable problem (e.g., the impact of hospital quality metrics on spending), explore it by analyzing the dataset (e.g., an observational study comparing high- vs. low-quality hospitals, or an interpretable ML model), and create an actionable recommendation.
What's the time commitment look like?
It's flexible and depends on your group and project.
Questions? Reach the organizers
- Emily Leventhal
- Michael Yao
- Lawrence Huang
Other editions