Skip to main content
Data / ML

Using Causal Inference to Improve the Uber User Experience

19 June 2019 / Global
Featured image for Using Causal Inference to Improve the Uber User Experience
Figure 1. Causal inference methods apply to very specific experimental data.
Figure 2. Causal inference methods apply to very specific observational data.
Figure 3. A possible data-generating causal graph shows how the number of Uber Eats orders could confound the relationship between experiencing a delayed delivery and customer engagement.
Figure 4. The regression discontinuity approach estimates if changing surge multiplier has an effect of purchase rates.21
Figure 5. A possible data-generating causal graph shows how an instrument is related with the candidate cause variable whose relationship with the outcome is confounded.
Totte Harinen

Totte Harinen

Totte Harinen is a senior data scientist with Uber Labs, Uber's Applied Behavioral Science team.

Bonnie Li

Bonnie Li

Bonnie Li is a senior data scientist with Uber Labs, Uber's Applied Behavioral Science team.

Posted by Totte Harinen, Bonnie Li

Category: