Spatio-Temporal Cab Demand Forecasting using Hybrid LSTM-CNN Models

Abstract

The research is aimed at designing and implementing an intelligent cab demand forecasting model that utilizes the methods of Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) models so that future ride demand can be accurately forecasted in an urban environment. Two groups were studied in this comparative study. Is the current state of the demand-forecasting system which is based on conventional statistical time-series analytics and machine-learning models, trained on a dataset of 1,000 past taxi-demand examples. The proposed system uses deep-learning to perform both spatial feature extraction with convolutional neural networks and temporal modeling with long short-term memory networks using the same number of observations in order to have a fair comparison. The suggested CNN-LSTM model demonstrated better predictive power and achieved an average accuracy of 91.4 in the study, and its root-mean-square error was lower than that of the existing methodology, which justifies its increased quality of demand forecasting. The experiment proves that the combination of convolutional neural networks and long short-term memory models significantly enhances the quality of cab demand predictions, which will result in the better organization of the fleet and will lead to the creation of intelligent transportation networks.