Taxi Demand Research Articles

Effect of Spatio-Temporal Granularity on Demand Prediction for Deep Learning Models

Abstract Advances in machine learning technology and the availability of big data from GPS systems have led to the development of effective methods for modelling transportation demand and forecasting the future. Most previous research concentrated on demand prediction using a variety of machine learning and deep learning models that took into account spatial and temporal relationships. This paper investigates the impact of spaces and time granularity for a Spatio-temporal demand modelling framework. Using taxi demand data from New York City, our study compares the prediction performance of deep learning models such as Long Short-Term Memory (LSTM), Convolution Neural Networks (CNN) and Temporal-Guided Networks (TGNet), modelled with a grid-based tessellation strategy. The findings of this study could assist researchers in better understanding how the granularity of space and time helps deep learning models perform better for demand forecasting problems.

Learning Spatial-Temporal Features of Ride-Hailing Services with Fusion Convolutional Networks

Accurate and reliable taxi demand prediction is of great importance for intelligent planning and management in the transportation system. To collectively forecast the taxi demand in all regions of a city, many existing studies focus on the capturing of spatial and temporal correlations among regions but ignore the local statistical differences throughout the geographical layout of a city. This limits the further improvement of prediction accuracy. In this paper, we propose a new deep learning framework, called the locally connected spatial-temporal fully convolutional neural network ( LC-ST-FCN), to learn the spatial-temporal correlations and local statistical differences among regions simultaneously. We evaluate the proposed model on a real dataset from a ride-hailing service platform (DiDi Chuxing) and observe significant improvements compared with a bunch of baseline models. Besides, we further explore the working mechanism of the proposed model by visualizing its feature extraction processes. The visualization results showed that our approach can better localize and capture useful features from spatial-related regions.

IMPACTS OF THE COVID-19 PANDEMIC ON DEMAND OF PUBLIC BIKE SYSTEM AND ITS CORRELATION WITH URBAN LIFE

Abstract. Public bike systems provide the flexibility of bike-usage that users can rent and return a bike freely at any station. The convenience of the bike-travel system, however, may turn into a disadvantage of demand-supply imbalance in the bike inventory. The recent spread of the COVID-19 pandemic has changed the mobility demands due to the lockdowns which restrict the business operating hours and transit services. Therefore, investigating the impacts of the pandemic on the urban social life patterns with the bike usage is an important issue for the bike demand prediction and the improvement of the public bike system service.This research aims to investigate the correlation between the public bike demand and social environment factors during the pandemic applying a multivariate linear regression model to public bike usage data in Seoul, Korea. The results show some promising findings to further promote shared mobility services through policy and marketing strategies. It is noteworthy that the transport disruptions during the pandemic have made a spillover effect from taxi and public transit to bike as an alternative transport mode. The lockdown has restricted the range of activity and resulted in the decrease of the taxi demand, so the number of taxis. On the other hand, the correlations of the geography, meteorology, and date with the bike demand have shown consistency. Therefore, supply of extra bike facilities to improve the system service should be determined based on more accurate demand prediction considering lifecycle-related factors.

Application of an Adaptive Adjacency Matrix-Based Graph Convolutional Neural Network in Taxi Demand Forecasting

Accurate forecasting of taxi demand has facilitated the rational allocation of urban public transport resources, reduced congestion in urban transport networks, and shortened passenger waiting time. However, virtual station discovery and modelling of the demand when forecasting through graph convolutional neural networks remains challenging. In this study, the virtual station discovery problem was addressed by using a two-stage clustering approach, which considers the geographical and load characteristics of taxi demand. Furthermore, a fusion model combining non-negative matrix decomposition and a graph convolutional neural network was proposed in order to extract the features of the nodes for dimension reduction and adaptive adjacency matrix computation. By the construction of a local processing structure, further extraction of the local characteristics of the demand was achieved. The experimental results show that the method in this study outperforms state-of-the-art methods in terms of the root mean square error and average absolute value error. Therefore, the model proposed in this study is able to achieve accurate forecasting of taxi demand.

Uncovering Factors Affecting Taxi Income from GPS Traces at the Directional Road Segment Level

Nowadays, the market demand for taxis is still intense. However, there exist lots of issues affecting the healthy development of the taxi industry, such as an increasing difficulty in hailing taxis, detouring behavior etc., and especially, the low incomes of taxi drivers. This paper establishes a multi-layer road index (MRI) system of 7862 directional road segments (DRSs), and collects over 194 million occupied GPS points within a week, revealing the factors affecting taxi drivers’ incomes in Shenzhen, China. The income differences has been identified on different DRSs, which accordingly have been categorized into two levels. Four categories of DRS factors, i.e., road attributes, traffic dynamics, points of interest (POIs), and taxi operation strategies, are defined as the impact factors affecting income levels. The selected sample-based binomial logit (SBL) model has been proposed to reveal the significance of these influencing factors. The results indicate that the road segments with different features have different incomes over different time periods. The main factors in income analysis are the factors used to represent taxi operation strategies. Highly rewarding pick-up road segments can be identified, which could contribute to drivers’ income improvements, and can further contribute to the development of the taxi market.

Modelling multiple quantiles together with the mean based on SA‐ConvLSTM for taxi pick‐up prediction

A more complete taxi pick-ups prediction going beyond the conditional expectation would be more beneficial to allocate taxis effectively. Prior works have predicted the conditional expectation of taxi demand and improved the prediction accuracy. But the ability to convey calibrated uncertainty estimates in prediction was less studied. Here, a multi-output multi-quantile deep neural network architecture is presented for predicting the mean and the multiple quantiles simultaneously, which jointly models multiple quantiles together with the conditional expectation based on self-attention ConvLSTM. Using the NYC taxi dataset, we have analyzed the performance of the developed architecture. For the mean prediction, the proposed method has an error of 8.584 in terms of RMSE criteria and Linear QR, Mcdropout-Gal, ConvLSTM and SA-ConvLSTM methods, which have error values of 8.800, 8.671, 8.599, and 8.597, respectively. According to the MAE criterion, this method has a value of 5.934. The other methods have obtained values of 6.614, 5.990, 5.986, and 5.959, respectively. For multiple quantiles prediction, the proposed method produces narrower intervals while obtaining the target coverage percentage. Overall, the proposed model not only outperforms the other models, but also provides a rich description about the predictive density.

Taxi‐demand forecasting using dynamic spatiotemporal analysis

Taxi-demand forecasting and hotspot prediction can be critical in reducing response times and designing a cost effective online taxi-booking model. Taxi demand in a region can be predicted by considering the past demand accumulated in that region over a span of time. However, other covariates—like neighborhood influence, sociodemographic parameters, and point-of-interest data—may also influence the spatiotemporal variation of demand. To study the effects of these covariates, in this paper, we propose three models that consider different covariates in order to select a set of independent variables. These models predict taxi demand in spatial units for a given temporal resolution using linear and ensemble regression. We eventually combine the characteristics (covariates) of each of these models to propose a robust forecasting framework which we call the combined covariates model (CCM). Experimental results show that the CCM performs better than the other models proposed in this paper.

BERT-Based Deep Spatial-Temporal Network for Taxi Demand Prediction

Taxi demand prediction plays a significant role in assisting the pre-allocation of taxi resources to avoid mismatches between demand and service, particularly in the era of the sharing economy and autonomous driving. However, most studies have only tried to figure out the complex spatial-temporal pattern of taxi demand from historical taxi demand series, neglecting the intrinsic influences of regional functions, and failing to effectively capture the dynamic long-term periodicity. In this paper, we make two important observations: (1) taxi demand pattern varies significantly between different functional regions; and (2) taxi demand follows a dynamic daily and weekly pattern. To address these two issues, we adopt Points of Interest (POIs) to identify regional functions, and propose a novel BERT-based Deep Spatial-Temporal Network (BDSTN) to model the complex spatial-temporal relations from heterogeneous local and global features. In BDSTN, a Spatiotemporal Pattern Matching module is introduced to capture the complex spatiotemporal pattern of taxi demand while considering its dynamic temporal periodicity, and a Functional Similarity Embedding module is adopted to learn the functional similarity among all regions via POIs. To the best of our knowledge, this is the first work to use BERT-based architecture to learn taxi demand patterns, and is also the first to take functional similarity represented by POIs into consideration. Our experimental results on real-world traffic datasets in New York City demonstrate that the effectiveness of the proposed method outperforms the state-of-the-art methods, and that the efficiency of our proposed model is higher than other deep learning methods.

Investigating the increasing demand and formal regulation of motorcycle taxis in Ghana

The demand for motorcycle taxis has seen a significant increase in Ghana in the last decade. In light of an existing unimplemented ban, several questions arise on the prospects of strong formal regulation and policy implementation. However, the lack of succinct research data on motorcycle taxi regulation from users' perspectives leaves a gap in policy formulation and makes it difficult for effective decision-making. In this study, a users' perspective approach is adopted to identify plausible factors that may inform policy direction for motorcycle taxi regulation in Ghana. Overall, the views of 1489 motorcycle taxi users were sampled. The analyses comprised correlations, multiple correspondence analysis, chi-square tests, and ordinal logistic regression. The results showed that 78% of the respondents support formal regulation but decentralized to the local authorities. Consequently, 70% of the respondents confirmed continuous use intentions. It is found that users' geographical location and occupations are significantly associated with continuous use intentions. Besides, nine influencing factors were confirmed to have a significant positive association with continuous use intentions. However, respondents' perceived safety of motorcycle taxis has a significant negative association with both support for regulation and continuous use intention. This study provides a novel framework, that is the Driving forces, Pressures, State, Impacts, Responses Framework (DPSIR), for the case of motorcycle taxi regulation in Ghana. The findings are relevant for transport policy and urban planning in Ghana and other countries of sub-Saharan Africa.

Taxi Demand Prediction using ML

Abstract: Taxi plays a crucial role in transportation especially in urban areas.Predicting the future demand for taxis in particular geographical location will greatly help internet based transportation companies like Ola, Uber etc. So that we can drastically decrease the waiting time of customers/passengers and also it helps taxi drivers to move to particular location where demand is high eventually making passengers,drivers and companies happy. In this Project we like to predict the demand for taxi in particular location for next 10 min using previous time series data .we want to perform this task of regression using machine learning models with high accuracy and then we would like to apply deep learning models and compare the results.we like to propose the best suited and high accuracy model for the problem.It will greatly help companies in managing the taxi fleet in cities.

Dynamic Stochastic Matching Under Limited Time

In “Dynamic Stochastic Matching Under Limited Time,” Aouad and Sarıtaç analyze the design of matching policies in dynamic markets such as carpooling platforms and kidney exchange schemes. A crucial distinction with previous literature is that the agents’ arrivals and departures are fully dynamic. The demand and supply side are constantly replenished; each market participant remains available for potential matches during a limited period of time. Specifically, the authors formulate a general dynamic matching model over edge-weighted graphs, where the agents' arrivals and abandonments are stochastic and heterogeneous. The platform controls how long each agent waits and whom s/he is matched with. These decisions are subject to a fundamental tradeoff between increasing market thickness and mitigating the risk of abandonments from certain participants. The authors’ main contribution is to devise simple matching algorithms with strong performance guarantees for a broad class of networks. In contrast, they show that widely used batching algorithms have an arbitrary bad performance on certain graph-theoretic structures. Their analysis involves novel techniques including linear programming benchmarks, value function approximations, and proxies for continuous-time Markov chains, which may be of broader interest. Extensive simulations on real-world taxi demand data demonstrate that the newly developed algorithms can significantly improve cost efficiency against batching algorithms.

Taxi Demand Forecasting Utilizing Neighbourhood Influence Based on Ensemble Learning

Abstract: This Time series forecasting (TSF) assists in making better strategic decisions under uncertain circumstances so that financial crisis can be avoided, wise investments can be made, under/over contracting of utility can be avoided, staffs can be scheduled appropriately, service providers can provide better service, mankind can get prepared for natural disasters and many more. However, the accuracy in forecasting plays a vital role and achieving such is a challenging task owing to the vagueness and nonlinearity associated with most of the real world time series. Therefore, improving the forecasting accuracy has become a keen area of interest among the forecasters from different domains of science and engineering. In this work, a time series forecasting model for Chicgo taxi trip data is proposed. The proposed model utilizes neighbourhood influence in terms of taxi demand to predict taxi demand. Keywords: Time series forecasting, Regularization, Regression

Exploring the Spatiotemporal Impacts of the Built Environment on Taxi Ridership Using Multisource Data

Taxis are an important component of the urban public transportation system, with wide geographical coverage and on-demand services characteristics. Thorough understanding of the built environment affecting taxi ridership can enable transportation authorities to develop targeted policies for transportation planning. Previous studies in this field had few data sources and did not consider the spatiotemporal variability. This study aims to develop an analytical framework for understanding the spatiotemporal correlation between the urban built environment and taxi ridership, which is empirically analyzed in New York City. The built environment is defined through multisource data in terms of density, design, diversity, and destination accessibility. Besides the exploration of travel patterns, the spatiotemporal heterogeneity of taxi ridership is modeled using geographically and temporally weighted regression (GTWR). The result shows that GTWR outperforms ordinary least squares (OLS), geographically weighted regression (GWR), and temporally weighted regression (TWR) in both goodness of fit and explanatory accuracy. More importantly, our study found that land use diversity is negatively correlated with taxi ridership, while transportation diversity is positively correlated with it. A highly accessible road network improves the people’s demand for taxis in the morning rush hours. Moreover, the density of railway stations is positively correlated with taxi ridership on weekdays but adversely on weekends. These findings provide practical insights for urban transportation policy development and taxicab regulation.

Spatio‐temporal neural network for taxi demand prediction using multisource urban data

AbstractPredicting traffic demand is a crucial problem to saving energy and resource pre‐allocation as well as traffic construction planning. The problem is also challenging owing to its non‐linear changing process and complex spatio‐temporal dependencies. Most existing studies train prediction models on traffic data itself and ignore the impact of exogenous factors, thus leading to unsatisfactory performance. To address this issue, we propose a deep neural network, which utilizes multisource urban data to capture specific geospatial and temporal dependencies, for city‐scale taxi demand forecasting. In terms of time dependency, a temporal guidance block is developed to enhance the correlation between time intervals by incorporating time metadata. To capture the geospatial dependencies, we build three parallel geospatial guidance branches that integrate multiple urban data, including population data, POIs, and road networks, to guide the modeling. Each branch contains a geocontext‐aware convolutional layer for extracting geospatial features and a self‐attention mechanism for capturing long‐range spatial dependencies. Extensive experiments have been conducted to demonstrate the superiority of our model over the existing methods, and the prediction error is 16% lower than the state‐of‐the‐art model. We also explore the sensitivity to its parameters and discuss the contributions and limitations of the proposed method.

The effects of peak hour and congested area taxi surcharges on customers’ travel decisions: Empirical evidence and policy implications

Under a stationary taxi fare structure, the varying passenger demand for taxis over time and space creates a serious shortage of taxi supply during peak hours and outside the city center. Apart from providing additional supply by increasing taxi fleet size and suppressing overall passenger demand by increasing taxi fare, it is more preferable to address the problem by means of surcharge. A surcharge can be imposed on the taxi customers who take taxis during peak hours and/or travel towards congested areas. A portion of the taxi customers may change their mode choices, arrival times, and taxi drop-off locations to avoid the additional cost. By alleviating the shortage of taxi supply, waiting time can be greatly reduced and thus the service quality can also be improved. In this study, 773 taxi customers were interviewed for their travel decisions under different hypothetical scenarios. A total of 4638 observations were used to calibrate multinomial logit and nested logit models for the analysis. The results of the logit models demonstrate that all six concerning attributes, walk and wait time, taxi travel time, public transport travel time, travel fare, early arrival time, and the number of interchanges involved, are significant to affect the taxi customers’ travel decisions. The results of market segmentation analysis further show the variations in the travel decisions of taxi customers in different segments, including gender, occupation, monthly income, and family car ownership. Sensitivity analyses were carried out, and suggestions on the implementation of a taxi surcharge scheme are provided. The calibrated models, findings, and discussions are believed to be useful for establishing a new taxi fare structure to strike a balance of spatio-temporal taxi demand and supply in the market.

BP Neural Network-Based Big Data Intelligent Travel Algorithm and Its Application

With the increase of urbanization rate, a large number of people flood into cities, increasing pressure of urban traffic, and problems accumulated in the taxi industry are gradually prominent. The phenomenon of crowded queue for taxi is frequent in peak hours, and vehicles patrol and sweep streets during peak hours. The key to solve these problems lies in mastering the rules and patterns of taxi travel and finding the factors affecting the relationship between taxi supply and demand. It is difficult to effectively understand taxi travel as a whole due to the large number of taxis and their large scale and strong mobility. Comprehensive application of trajectory data mining method can extract the spatiotemporal characteristics of massive taxi trajectory data and reveal the nature of its occurrence. This paper mainly focuses on the problems faced by urban traffic governance, such as the mismatch between data sources and demand systems, the uncoordinated operation of comprehensive transportation system, and the difficulty in sharing big data resources between government and enterprises. With massive data resources across departments, this paper designed a networked intelligent computing platform for big data of urban transportation integrated with various modes of transportation to sense the operation situation of urban comprehensive transportation system in real time, accurately grasp the space-time distribution of urban transportation supply and demand, and significantly improve the ability of coordinated operation, organization, and management of transportation in large cities. It also effectively enhances the quality of transportation information sharing and integration services and comprehensively improves the efficiency and overall carrying capacity of the urban comprehensive transportation system.

Spatial-Temporal Diffusion Convolutional Network: A Novel Framework for Taxi Demand Forecasting

Taxi demand forecasting plays an important role in ride-hailing services. Accurate taxi demand forecasting can assist taxi companies in pre-allocating taxis, improving vehicle utilization, reducing waiting time, and alleviating traffic congestion. It is a challenging task due to the highly non-linear and complicated spatial-temporal patterns of the taxi data. Most of the existing taxi demand forecasting methods lack the ability to capture the dynamic spatial-temporal dependencies among regions. They either fail to consider the limitations of Graph Neural Networks or do not efficiently capture the long-term temporal dependencies. In this paper, we propose a Spatial-Temporal Diffusion Convolutional Network (ST-DCN) for taxi demand forecasting. The dynamic spatial dependencies are efficiently captured through a two-phase graph diffusion convolutional network where the attention mechanism is introduced. Moreover, a novel temporal convolution module is designed to learn various ranges of temporal dependencies, including recent, daily, and weekly periods. Inside the module, convolution layers are stacked to handle very long sequences. Experimental results on two large-scale real-world taxi datasets from New York City (NYC) and Chengdu demonstrate that our method significantly outperforms seven state-of-the-art baseline methods.

A multi-sensory stimulating attention model for cities\u2019 taxi service demand prediction

Taxi demand forecasting is crucial to building an efficient transportation system in a smart city. Accurate taxi demand forecasting could help the taxi management platform to allocate taxi resources in advance, alleviate traffic congestion, and reduce passenger waiting time. Thus, more efforts in industrial and academic circles have been directed towards the cities’ taxi service demand prediction (CTSDP). However, the complex nonlinear spatio-temporal relationship in demand data makes it challenging to construct an accurate forecasting model. There remain challenges in perceiving the micro spatial characteristics and the macro periodicity characteristics from cities’ taxi service demand data. What’s more, the existing methods are significantly insufficient for exploring the potential multi-time patterns from these demand data. To meet the above challenges, and also stimulated by the human perception mechanism, we propose a Multi-Sensory Stimulus Attention (MSSA) model for CTSDP. Specifically, the MSSA model integrates a detail perception attention and a stimulus variety attention for capturing the micro and macro characteristics from massive historical demand data, respectively. The multiple time resolution modules are employed to capture multiple potential spatio-temporal periodic features from massive historical demand data. Extensive experiments on the yellow taxi trip records data in Manhattan show that the MSSA model outperforms the state-of-the-art baselines.

A Decision Framework to Recommend Cruising Locations for Taxi Drivers under the Constraint of Booking Information

As the demand for taxi reservation services has increased, increasing the income of taxi drivers with advanced services has attracted attention. In this article, we propose a path decision framework that considers real-time spatial-temporal predictions and traffic network information. The goal is to optimize a taxi driver's profit when considering a reservation. Our framework contains four components. First, we build a grid-based road network graph for modeling traffic network information for speeding up the search process. Next, we conduct two prediction modules that adopt advanced deep learning techniques to guide proper search directions for recommending cruising locations. One module of the taxi demand prediction is used to estimate the pick-up probabilities of passengers in the city. Another one is destination prediction, which can predict the distribution of drop-off probabilities and capture the flow of potential passengers. Finally, we propose the H* (Heuristic-star) algorithm, which jointly considers pick-up probabilities, drop-off distribution, road network, distance, and time factors based on the attentive heuristic function to dynamically recommend next cruising locations. Compared with existing route planning methods, the experimental results on a real-world dataset have shown that our proposed approach is more effective and robust. Moreover, our designed search scheme in H* can decrease the computing time and allow the search process to be more efficient. To the best of our knowledge, this is the first work that focuses on guiding a route, which can increase the income of taxi drivers under the constraint of booking information.

Taxi Demand Prediction Using Parallel Multi-Task Learning Model

Accurate and real-time taxi demand prediction can help managers pre-allocate taxi resources in cities, which assists drivers quickly finding passengers and reduce passengers’ waiting time. Most of the existing studies focus on mining spatial-temporal characteristics of taxi demand distributions, while lacking in modeling the correlations between taxi pick-up demand and the drop-off demand from the perspective of multi-task learning. In this article, we propose a multi-task learning model containing three parallel LSTM layers to co-predict taxi pick-up and drop-off demands, and compare the performance of single demand prediction methodology and that of two demands’ co-prediction methodology. Experimental results on real-world datasets demonstrate that the pick-up demand and the drop-off demand do depend on each other, and the effectiveness of the proposed co-prediction methods.