Data Science Hacks - Google Causal Impact
Inferring Causal Impact with Google's Causal Impact Package
What you will learn
Inferring the Causal Impact of a event (Promotion, Marketing Campaign, etc) over sales, website visits, download apps or any other variable you want to analyse
How to use the pythons implementation of Google Causal Impact package
How to calculate the ROI of a marketing campaign or a sales promotion
Why take this course?
Welcome to our Google Causal Impact Course.
This course I'll teach you how to use the google's package Causal Impact in your on job or personal projects.
The Causal Impact model developed by Google works by fitting a bayesian structural time series model to observed data which is later used for predicting what the results would be had no intervention happened in a given time period. The idea is to used the predictions of the fitted model (depicted in blue) as a reference to what probably would had been observed with no intervention taking place.
After this course you will have a powerful tool, to measure (with statistical significance):
* The extra number of sales / app downloads / clicks / web site visits caused by a marketing campaign
* The ROI of a Marketing Campaign
* The effect of a promotion over demand
* Any change of behavior in a series, caused by a known event
Content
Course Introduction
Model Inputs
Building the Model
Interpreting the Results
Screenshots
Reviews
Charts
Price
Rating
Enrollment distribution
