Abstract:In spite of its importance, passenger demand prediction is a highly challenging problem, because the demand is simultaneously influenced by the complex interactions among many spatial and temporal factors and other external factors such as weather. To address this problem, we propose a Spatio-TEmporal Fuzzy neural Network (STEF-Net) to accurately predict passenger demands incorporating the complex interactions of all known important factors. We design an end-to-end learning framework with different neural networks modeling different factors. Specifically, we propose to capture spatio-temporal feature interactions via a convolutional long short-term memory network and model external factors via a fuzzy neural network that handles data uncertainty significantly better than deterministic methods. To keep the temporal relations when fusing two networks and emphasize discriminative spatio-temporal feature interactions, we employ a novel feature fusion method with a convolution operation and an attention layer. As far as we know, our work is the first to fuse a deep recurrent neural network and a fuzzy neural network to model complex spatial-temporal feature interactions with additional uncertain input features for predictive learning. Experiments on a large-scale real-world dataset show that our model achieves more than 10% improvement over the state-of-the-art approaches.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the complexity and challenges in **passenger demand prediction**. Specifically, passenger demand is comprehensively affected by a variety of spatial, temporal and external factors, including but not limited to weather, time period (day or night), geographical location and time - series information. Accurately predicting future passenger demand is crucial for transportation network companies (TNCs) to optimize resource allocation, reduce passenger waiting time, decrease energy consumption and relieve traffic congestion. However, the existing methods have the following deficiencies in dealing with this problem: 1. **Traditional methods (such as ARIMA)** only consider time - series features and ignore spatial features. 2. **Early deep - learning models** can capture spatial and temporal features simultaneously, but usually process these features sequentially, resulting in information loss. 3. **Existing models fail to fully consider the influence of external factors (such as weather)** and do not handle the inaccuracy of external data well. To solve these problems, the author proposes a new deep spatio - temporal fuzzy neural network (Spatio - TEmporal Fuzzy neural Network, STEF - Net), aiming to predict passenger demand in urban areas more accurately. The main contributions of STEF - Net include: 1. **Integrating fuzzy neural network and deep - learning techniques** to handle data uncertainty and learn the complex interaction relationships between multiple factors. 2. **Proposing a new feature - fusion method** that uses convolution operations to connect two independent networks, retaining temporal information and capturing spatial information. 3. **Adopting an attention mechanism** to provide interpretable results at each time step, explaining how historical information affects the prediction and how the weather affects the prediction. 4. **Verified by a large number of real - data experiments**, it is proved that this model is significantly superior to the existing state - of - the - art models, with an improvement of more than 10% in prediction accuracy. ### Overview of the model structure STEF - Net mainly consists of four parts: 1. **ConvLSTM layer**: It is used to capture the spatio - temporal feature interaction. The input is the historical passenger demand with location and time information, and the output is the prediction result using only spatio - temporal information. 2. **Fuzzy neural network**: It is used to model the influence of external information on passenger demand. The input is external information data, and the output is the prediction result based on external information. 3. **Feature - fusion module**: The outputs from ConvLSTM and the fuzzy neural network are fused together through convolution operations, and the attention mechanism is further used to capture the temporal correlation. 4. **Loss function**: It is used to measure the difference between the predicted value and the real value, and the goal is to minimize the mean - square error (MSE). ### Conclusion By introducing the fuzzy neural network and the innovative feature - fusion method, STEF - Net can effectively deal with the uncertainty and complexity of external data while processing spatio - temporal features, thus achieving more accurate passenger demand prediction.

A Deep Spatio-Temporal Fuzzy Neural Network for Passenger Demand Prediction

Space-Time Adaptive Network for Origin-Destination Passenger Demand Prediction

Demand Forecasting and Predictability Identification of Ride-Sourcing Via Bidirectional Spatial-Temporal Transformer Neural Processes

Short-Term Forecasting of Passenger Demand under On-Demand Ride Services: A Spatio-Temporal Deep Learning Approach

Spatio-Temporal Graph Convolutional and Recurrent Networks for Citywide Passenger Demand Prediction

Passenger Demand Forecasting with Multi-Task Convolutional Recurrent Neural Networks.

Spatio-Temporal Dynamic Multi-graph Attention Network for Ride-Hailing Demand Prediction

Deep Multi-View Spatial-Temporal Network for Taxi Demand Prediction

Deep Learning for Metro Short-Term Origin-Destination Passenger Flow Forecasting Considering Section Capacity Utilization Ratio

DeepSTCL: A Deep Spatio-temporal ConvLSTM for Travel Demand Prediction

Transformer Based Spatial-Temporal Fusion Network for Metro Passenger Flow Forecasting.

Spatial-temporal multi-factor fusion graph neural network for traffic prediction

Spatial-Temporal Attention Fusion Network for short-term passenger flow prediction on holidays in urban rail transit systems

Deep spatio-temporal neural network based on interactive attention for traffic flow prediction

Predicting Ride-Hailing Passenger Demand: A POI-based Adaptive Clustering Federated Learning Approach

Short-term Passenger Flow Forecasting of the Airport Based on Deep Learning Spatial-temporal Network

Short-Term Origin-Destination Demand Prediction Based on Spatiotemporal Encoder-Decoder Network with a Residual Feature Extractor

Multi-sequence spatio-temporal feature fusion network for peak-hour passenger flow prediction in urban rail transit

Short-Term Online Forecasting for Passenger Origin–Destination (OD) Flows of Urban Rail Transit: A Graph–Temporal Fused Deep Learning Method

A network traffic prediction model based on reinforced staged feature interaction and fusion

A multi-feature spatial–temporal fusion network for traffic flow prediction