It is important to predict the final destination of a taxi to enhance the efficiency of electronic taxi dispatching systems. When there is high demand, often there could be a taxi whose current ride can end near or exactly at the requested pick up location from a new rider. Predicting the final destination of taxi will help the dispatcher know where the driver would be ending their current ride so they can identify which taxi should be assigned for the next pickup request.
In this data science project, you will predict the destination of a taxi given the variable-length sequence of GPS points which represent the beginnig of its trajectory and other information in the taxi trajectory dataset such as taxi id, client information, date and time. The training data has details of all the taxi rides from 2013-2014 in the city of Porto, Portugal. The taxi trajectory dataset represent around 1.7 million rides run by 442 taxis.