Abtest and Information Geometry

Happy new year!

Like I said in the last post, I joined meta in Aug. 2022. In meta my work is all about AB test (most with signal loss, i.e., conversion/signal data is not available at some granularity levels). This leads a new idea: can we use information geometry to deliver some new AB test framework?

Here are some references I collected from the google scholar:

Information Geometry of Statistical Inference - An Overview, Shun-ichi Amari

Information geometry in optimization, machine learning and statistical inference, Shun-ichi AMARI

Information geometry in cosmological inference problems, Eileen Giesel etc.

Hypothesis testing, information divergence and computational geometry, Frank Nielsen

Information Geometry, Nihat Ay etc. (book)

Parameter estimation and uncertainty quantification using information geometry, Jesse A. Sharp etc.

Let’s see if I can find any insight from them.