In October 2018, I was invited as a mentor to attend Health IoT Hackathon in Taipei. This is a four-day event co-hosted by Taipei Medical University (TMU) and Massachusetts Institute of Technology (MIT). Students in medical school, data science, and engineering school teamed up with practitioners in hospitals to explore novel applied data science projects and digital product prototyping.
Within four days, each team will conceptualize a meaningful question, explore available data sources, develop computation methods with statistical modeling and machine learning, validate the results, and prepare for a presentation. Here are my two major takeaways:
- In a hackathon, defining a problem and exploring data should be an iterative, ideally concurrent process. In short, we need need to find the right data to answer a meaningful question. However, it is questionable whether hackathon forces us to seek “low-hanging fruits”, simply due to its limit time.
- During the early stage, it would be beneficial for the team to reach a consensus on the value proposition of their project. Specifically, what are the expected deliverable and its potential impact/ importance/ value? This seems to be a bottleneck for many teams that they do not have a vision of the final product (scientific discovery? hidden problem? algorithm? product?) even after defining a valid question.
I believe these are common challenges not only for people in hackathons but in general data-driven research as well. These challenges worth future studies to better understand the research management strategies on interdisciplinary data project and prototyping. Finally, I would like to thank MIT team for having me on this event. I have learned so much from other researchers and glad to contribute my feedbacks to support the clinicians and data scientists in Taipei.