0
$\begingroup$

I have a dataset which consists of calls placed to 911 occurring over a period of 13 years. The annual total calls has increased significantly over time, starting at around 200000 per year, and then increasing to at least 500000 after the first 3 years. The total for the most recent 5 year period is closer to 700000 annually.

The data consists of a unique event identifier and a time stamp - as well as other details.

My goal is to identify “surges” of calls occurring within several separate fixed time periods: 1 minute, 2 minutes, 5 minutes and 10 minutes.

Improve this question
$\endgroup$
1
  • $\begingroup$ Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. $\endgroup$
    – Community Bot
    Commented Mar 17, 2022 at 12:14

1 Answer 1

Reset to default
0
$\begingroup$

A potential method is performing change point detection. However, within the field of change point detection, there exists a wide variety of techniques applicable to particular problem statements.

In your case, it seems you would like to perform a "sliding fixed window" (1, 2, 5, 10 minute) change point. A very basic algorithm would be to compute an aggregate statistic (max, min, average number of calls) for each time window. More advanced algorithms detect a shift in the mean or variance for a time window. Which, in your case, should be detected by a change in mean over longer time period windows such as year(s).

If you use Python, the following Ruptures package offers some of the more advanced change point detection methods.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.