“ Study the past if you would define the future. “ — Confucius
If you want to study the past then GIDEON is a great place to do so. The question is, what would you do with over 750,000 data points, dating back to the 17th Century?
A team of researchers answered in a recent interview.
- Tad Dallas, Assistant Professor at Louisiana State University
- Colin J. Carlson, Assistant Research Professor at Georgetown University
- Timothée Poisot, Assistant Professor at Université de Montréal
“The three of us have talked a lot about why predictions and ecology work or don’t work. We all share a lot of frustrations about where things fall short and how things get framed. We were having a conversation one day about what sort of minimum information is that you would need to know what the future looks like. Then Tad said: “I’m going to make a model with no predictors”.
“It opened up this incredibly interesting question — What do we do with this model if it’s able to do predictions?”
How did the GIDEON data help with the study?
“We tried to make a completely predictor-free model to just see if we can let the similarity of pathogen communities to inform prediction and see how this can influence the future.”
“The main question was to get a sense of what happens where and when it happens”, said Colin. “There isn’t any cohesive narrative about where outbreaks happen. We used the GIDEON data because we thought it might be the most complete thing there is.”
The outcome
The team produced a research paper “ Testing predictability of disease outbreaks with a simple model of pathogen biogeography “, published by The Royal Society in November 2019. It proved effective at making the disease outbreak predictions and was created using only GIDEON data. Impressive!
The value of data is only realized when it is put into context and made relevant to a particular problem or theory. You provide the context and the problem, and we’ll provide you with the data. How could GIDEON fuel your research?
Originally published at https://www.gideononline.com.
