I have a time series dataset that describes events occurring over two years. For each event, I have metadata for every hour in the 24 hours around that event, and a categorical variable that describes the event. For example, for an event that occurs at 3 pm on Dec 12th, I have information about wind speeds and humidity for every hour in the 12 hours before and after that event, as well as how many days passed since the previous event and a categorical variable that describes the type of soil at that location.

I can also add a time series with the metadata where the event did not happen. Overall I have ~400 events.

I would like to find out the following:

  1. Create some form of a regression that can be used to predict if an event will occur by looking at the wind speeds and humidity over 12 hours before the event as well as how much time passed since the previous event.

  2. Find out parameters that are the most important for the prediction.

I was hoping you could point me to the correct models I should use and maybe to some methodologies on how to conduct the analysis in either Python/R/JMP.

Thank you !!



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy