I'm working on a Kaggle competition about predicting subsurface models - a map of the geologic layers beneath a given area - from seismic data.
I don't expect to place competitively but I learn a lot from these competitions. I like competitions like this that are connected to physical problems and datasets (though sadly this one is largely simulated), I learn as much about the broader world as I do deep learning. I've always idly wondered how seismology worked, and now I have an excuse to dig into it.
It's also given me a greater appreciation for the "seismology" we perform in our day to day, like knocking on things to see if they're hollow, or (as I learned here a while back) the way battitores test the porosity of cheese by knocking on it with hammers.
I don't expect to place competitively but I learn a lot from these competitions. I like competitions like this that are connected to physical problems and datasets (though sadly this one is largely simulated), I learn as much about the broader world as I do deep learning. I've always idly wondered how seismology worked, and now I have an excuse to dig into it.
It's also given me a greater appreciation for the "seismology" we perform in our day to day, like knocking on things to see if they're hollow, or (as I learned here a while back) the way battitores test the porosity of cheese by knocking on it with hammers.