A company has developed a new web page in order to increase the number of users who "convert" (the number of users who decide to pay for the company's product). For this project, I performed A/B tests to analyze the changes (conversion rate of the shoppers) through statical conclusions; implement the new page, keep the old page, or perhaps run the experiment longer to make their decision.
- Preprocessing
- Two-Proportion Z-Test
- A/B Testing
- Logistic Regression
- Python (Pandas, Numpy, Random, PyPlot, Scipy, Statsmodels API)
- ab_data.csv
Variable Name |
Metadata |
user_id |
6-digit numbers |
timestamp |
string |
group |
string: control, treatment |
landing_page |
string: old_page, new_page |
converted |
numeric: 0:No, 1:Yes |
- countries.csv
Variable Name |
Metadata |
user_id |
6-digit numbers |
country |
string: US, CA, UK |