building a model that calculate effect of an event

Say I have a bike sharing company and want to know how events affect number of rides per day.

events vary on:

  • number of coupons sent per user
  • expiry date
  • amount for each coupon.

I have created a df that looks something like this:

deploy  duration  avg_temp  rain  register_count   ...   rent_count
   33     131113      23      2.2       12                   33
   523    931143      25      0         322                  756
   63     231153      26      0         111                  124 
   123    566363      22      1.2       334                  345 

where rent_count is our target variable.

If I want to add columns that describe events(num_coupons, duration, etc...) how should I design columns?

So at the end I want to build a model that would look something like this:

Y = b0 + b1x1^n + .... + bnxn^n.

therefore I can calculate how much event variable is contributing to Y.

Topic cause-and-effect machine-learning-model machine-learning

Category Data Science


This is a classic example of Causality, where we need to find how much one variable is effected by changing other variable. You can infer the causal graph from data and then select the variables which best describe your model. Intervention can be done to represent particular event and then check the effect on target variable.

Refer: Causalnex , DoWhy

About

Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.