Please enable JavaScript.
Coggle requires JavaScript to display documents.
a/b testing (Distributions (Continuous, Discrete, joint, Conditional, PDF,…
a/b testing
Distributions
Continuous
Discrete
joint
Conditional
PDF
PMF
CDF
P-value
We should not finish test after getting significance value
It should be measured at the end of the test
It's the probability of detecting a difference even when 2 groups are the same
Underflow problem
Use the log probability to avoid
Probability -> 0, eventually computer will round to 0
A/B/C/D/... Testing
Check statistical significance
Pairwise testing
Post Hoc testing
One vs the rest
Statistical Power
Power Analysis
Can be used to determine the quantity of samples
Sample size calculator
Look for calculators online
Bayesian way
epsilon-greedy
Explore-exploit dilemma
Conversion Rate Optimization