I have time series of more then 6000 SKUs. I want to forecast for the future buckets. How to determine if a SKU time series has trend, seasonality or both ? Which technique suits which SKU ?
I strongly recommend Forecasting: Principles and Practice, 3rd edition:
Rob Hyndman created the forecast package and his team developed its successor, the tidyverts set of packages (fable, feasts, and tsibble).
Section 4.3 covers how to measure of the strength of trend and seasonality. A relevant quote from that section is "These measures can be useful, for example, when you have a large collection of time series, and you need to find the series with the most trend or the most seasonality."
To understand the tidyverts packages in general and these measures specifically, you should read the earlier chapters. If you are asking how to choose an appropriate forecasting method, you will want to read the entire book.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.