0

I'm looking for research related to the value of data volume as time decreases approaching an event.

For example, large data sets are needed to understand changes in geological and meteorological trends across centuries. A smaller data set is needed from satellites, radars, airborne sensors, and ground stations to predict weather patterns in a region for the next week. the weather outside your home can be predicted for the next few hours using only current temperature, humidity, barometric pressure, and wind speed. If you want to have a barbecue, the data volume required diminishes as the event approaches.

Hoping to find research that shows whether or not data volume becomes less valuable as time decreases before an event.

Matt E.
  • 1
  • 1
    Not sure there would be a general body of knowledge on this one. A simple explanation could be something along the line of the smaller the time period you have to predict in advance for, the simpler model you can get away with, and the simpler the model, the less data you need to fit it. – Cryo Jul 14 '23 at 23:21
  • 1
    In your example with weather. To predict far in advance you need to simulate weather system with millions of parameters. To predict few hours in advance you could probably get away with ARIMA type model – Cryo Jul 14 '23 at 23:23

0 Answers0