FP_Baier_Giometti_PE

[njgio]

Fabian,

Your idea is cool. Let me make sure I understand it correctly: Can we predict the magnitude of an earthquake based on the current weather? In your implementation you're specifically asking, based on the max temp of the day, can you predict how big or small an earth quake will be?

This could be a big hurdle to prove with statistical significance that a day's temp can affect the magnitude of that day's earthquake. At first glance this seems like it might be a correlation vs causation problem. As in, there could be a spurious correlation between number of times I click my pen and the percent of questions I get right on a test... But maybe there is some cool unexplored seismological phenomenon at play here.

An interesting thing to look at in terms EDA is where are most earthquakes occurring? Do earthquakes happen during a certain time of the year? Are they more likely in winter months? And if so, what is the average magnitude of the earthquake? Is it higher than the earthquakes during the summer time? If there is a correlation, then that begs some pretty cool questions like, does weather affect tectonic activity in general? Maybe if it's very cold, the water in aquifers will freeze, causing ground swell, which causes plates to move and build up pressure, causing a chain reaction that leads to a big ole quake.

Unfortunately, I could get a lot of code to work on my machine - something with the def weather(): function went haywire and I couldn't join max temp to the earthquake dataframe, so I can't really comment on the majority of your python work.

You might also consider making a binary classifier like: big quake, with a certain threshold for magnitude that would classify large or small earthquakes. Then you could use a whole bunch of the other ML techniques.

Anyway, pretty cool idea and there's lots of interesting work to be done. I'm excited to see how the project comes along.

Cheers,
Nick G