003 : Time series

Summary

This project develops diverse models to predict the number of taxi orders in the next hour at airports. The company Sweet Lift Taxi aims to create a tool to optimize driver availability during peak hours.
The main objective is to implement a model that anticipates taxi demand within a one-hour time window, based on historical order data from airports. To be effective, the model must achieve an RMSE metric no greater than 48 on the test set.

Github repository

Linear regression evaluation

Time series decomposition

Making features