2

I've got a bunch of data on pop singers' performance on the Hot 100 charts over time, and I'm trying to compare the early part of different artists' careers. For example, I might look compare Miley Cyrus' career to the commensurate period of Cher's career:

Time series plots of performance

(The y axis is total "top 10 points" --- 10 points for a weekly #1, etc.)

I'm trying to identify the singer among the 2000-some in the database who had the most similar early career. Running a simple correlation on the first 7 years of everyone's career wasn't very fruitful:

enter image description here

My next idea was to fit a binomial to each person and compare. Here's Beyonce vs. The Beatles through six years, for example:

enter image description here

The trouble is, I have no idea how to compare a bunch of polynomials to figure out where they're similar.

Any ideas how to do that, or otherwise how to compare these datasets?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Chris Wilson
  • 389
  • 1
  • 3
  • 12
  • What you're talking about is Time Series Analysis. https://stats.stackexchange.com/questions/19103/how-to-statistically-compare-two-time-series should give you the approach you need. – Blake Aug 20 '14 at 22:11
  • You could look into time series clustering https://stats.stackexchange.com/questions/3238/time-series-clustering-in-r, https://stats.stackexchange.com/questions/9475/time-series-clustering – kjetil b halvorsen May 09 '21 at 01:02

0 Answers0