Taxi Demand Prediction using Ensemble Model Based on RNNs and XGBOOST

Abstract

Taxis play an important role in urban transportation. Understanding the taxi demand in the future gives an opportunity to organize the taxi fleet better. It also reduces the waiting time of passengers and cruising time of taxi drivers. Even, there are some works proposed to predict the demand of taxi but there are few studies that consider the function of areas such as hospital area, department store area, residential area, and tourist attraction. One predictive model may not fit with all types of area. We use a point of interest (POI) to match taxi demand with a place to study the taxi demand in the area with a different function. In this paper, we investigate the best predictive models that can forecast demand of taxi hourly with 7 types of area function. The models that were selected for the experiment are long short term memory (LSTM), gated recurrent unit (GRU) and extreme gradient boosting (XGBOOST). Then, we proposed the ensemble model that can forecast the taxi demand well with all types of area function using the information from those machine learning models. We build the models based on a real-world dataset generated by over 5,000 taxis in Bangkok, Thailand for 4 months. The result shows that the proposed ensemble model can outperform other models in overall.