Machine learning for Taxi Business

Alternative Text
by Dmitry Elisov
Oqtacore CTO
226
alt

We did machine learing for Taxi Business. an analysis for a taxi agency based on their daily numbers, and found that one of the most important factors is how their customers are satisfied – the higher satisfaction.

How we did Machine Learning for Taxi Business

The owners of the taxi business already had a system of collecting key metrics:

  • Number of customer complaints in each city
  • Number of SMS orders in each city
  • Number of vehicles working in each city
  • Number of broken vehicles in each city
  • Number of new drivers in each city
  • Number of fired drivers in each city
  • Number of cancelled orders in each city

Our task was to use these numbers and present an analysis of the data with some simple conclusions.

We decided to use Random Forest algorithm for machine learning for Taxi business. We chose it because it is very easy to tell, compared to neural networks that cannot give conclusions that can be easily understood.

You can read about our other data analysis projects.

Taxi data analysis projects.

Rate this article
Please wait...