|
This article is cited in 1 scientific paper (total in 1 paper)
Methods for accelerating controlled online experiments
A. D. Minakov, V. A. Sudakov
Abstract:
The paper discusses current methods of hypothesis testing on a large set of data on user actions on a banking site. Methods have been developed and investigated using a consistent approach and Bayesian statistics, and a recommendation for use has been formed. The application of the proposed methods made it possible to reduce the time of conducting controlled online experiments by 25-34%.
Keywords:
testing, hypothesis testing, sequential testing, data analysis.
Citation:
A. D. Minakov, V. A. Sudakov, “Methods for accelerating controlled online experiments”, Keldysh Institute preprints, 2023, 036, 16 pp.
Linking options:
https://www.mathnet.ru/eng/ipmp3164 https://www.mathnet.ru/eng/ipmp/y2023/p36
|
Statistics & downloads: |
Abstract page: | 59 | Full-text PDF : | 14 | References: | 14 |
|