Documents DataMa
  • Home
  • Solutions
    • DataMa Compare
    • DataMa Pivot
    • DataMa Impact
  • News
  • About
  • Confidentiality Agreement
  • DataMa Compare
  • DataMa Impact
  • DataMa Journey
  • DataMa Pivot
  • DataMa Software License Agreement
  • Extensions
  • All Docs

DataMa Impact

  • General introduction
  • Web application
    • Introduction
    • Dashboards
      • Value boxes
      • Models
        • Frequentist (Bernouilli)
        • Frequentist (other)
        • Forecasting
        • Boxplot
        • Bayesian Bernouilli
    • Menu
      • Definitions
      • Filters
        • Excluded dimension
        • Filters n°1 & 2
      • Settings
        • Test to compute
        • Confidence interval
        • Aggregation in %
      • 2.3 Admin header
        • Loaded version
        • Change source
          • CSV
          • Excel
          • Google Analytics
          • Gsheet
        • Save as…
        • Manage version
    • Inputs
      • Sources
      • Metrics relation
    • Outputs
      • Shiny Web App
Home DataMa Impact Web application Dashboards Models Bayesian Bernouilli

Bayesian Bernouilli

In the bayesian Bernoulli approach, the outcome has to be binary (e.g. Purchase). As long as this is true, any aggregation of the data can be used.

Behind the scenes: A bayesian test is computed, based on an uninformative beta prior for the probability of success.

 

See also: https://www.evanmiller.org/bayesian-ab-testing.html

Test data: https://docs.google.com/spreadsheets/d/1VJJ2j5ldrSfvLQatd9SAikIJX_2dhBgDCjkdX_oUgB4/edit#gid=0

Updated on décembre 10, 2019
© 2021 Documents DataMa
Proudly powered by WordPress | Theme: weDocs by Tareq Hasan