Last year, ByteSumo published a detailed presentation explaining the detailed workings of the trendcaluclus method which describes how the calculations work, and some of the things we’re trialling in terms of using it as a feature generator for machine learning.
This was the first version of the document released to the general public as a download.
You can find it posted on our github page here:
https://github.com/bytesumo/TrendCalculus
