A Taxi Gap Prediction Method via Double Ensemble Gradient Boosting Decision Tree

Zhang, Xiao; Wang, Xiaorong; Chen, Wei; Tao, Jie; Huang, Weijing; Wang, Tengjiao

doi:10.1109/bigdatasecurity.2017.27

Cited by 15 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although many studies have focused on taxi/ridehailing demand forecasting problem, very few studies have explored the ride-hailing supply-demand gap forecasting problem. Some of the previous studies [29,30] utilized ensemble of various machine learning algorithms for forecasting the region-wise supply-demand gaps. [30] applied a multi-layer ensemble of various machine learning algorithms (e.g., support vector machine, single XGBoost, bagging XGBoost, random forest, extra trees, and AdaBoost).…”

Section: Related Workmentioning

confidence: 99%

Using spatio‐temporal deep learning for forecasting demand and supply‐demand gap in ride‐hailing system with anonymised spatial adjacency information

Rahman

Rifaat

2021

IET Intelligent Trans Sys

View full text Add to dashboard Cite

To reduce passenger waiting time and driver search friction, ride-hailing companies need to accurately forecast spatio-temporal demand and supply-demand gap. However, due to spatio-temporal dependencies pertaining to demand and supply-demand gap in a ridehailing system, making accurate forecasts for both demand and supply-demand gap is a difficult task. Furthermore, due to confidentiality and privacy issues, ride-hailing data are sometimes released to the researchers by removing spatial adjacency information of the zones, which hinders the detection of spatio-temporal dependencies. To that end, a novel spatio-temporal deep learning architecture is proposed in this paper for forecasting demand and supply-demand gap in a ride-hailing system with anonymised spatial adjacency information, which integrates feature importance layer with a spatio-temporal deep learning architecture containing 1D convolutional neural network (CNN) and zone-distributed independently recurrent neural network (IndRNN). The developed architecture is tested with real-world datasets of Didi Chuxing, which shows that the models based on the proposed architecture can outperform conventional time-series models (e.g. ARIMA) and machine learning models (e.g. gradient boosting machine, distributed random forest, generalized linear model, artificial neural network). Additionally, the feature importance layer provides an interpretation of the model by revealing the contribution of the input features utilized in prediction.

show abstract

Section: Related Workmentioning

confidence: 99%

Using spatio‐temporal deep learning for forecasting demand and supply‐demand gap in ride‐hailing system with anonymised spatial adjacency information

Rahman

Rifaat

2021

IET Intelligent Trans Sys

View full text Add to dashboard Cite

show abstract

“…malls and restaurants); more people will resort to OCH services in bad weather; traffic jam dampens the interest in OCH services. Many scholars [7][8][9][10] have designed excellent path planning models that recommend the best itinerary to drivers, yet failed to assess the contribution of POIs to OCH supply-demand. Chen et al [2] noticed the relationship between POIs and supply-demand gap, but did not clearly demonstrate the relationship.…”

Section: Literature Reviewmentioning

confidence: 99%

Supply-Demand Prediction of DiDi Based on Points of Interests Selection in Extreme Gradient Boosting Algorithm

Zheng¹,

Li²,

Zhang³

et al. 2020

RIA

View full text Add to dashboard Cite

In recent years, DiDi, an online car-hailing (OCH) service provider, has emerged as a leader in the sharing economy. To improve user experience, the company must minimize the waiting time and optimize car utilization based on accurate estimation of supply-demand gap. This paper aims to develop a desirable model to select the most significant factors for OCH supply-demand estimation. Firstly, the correlation between the points of interest (POIs) and the supply-demand gap was proved through statistical analysis. Next, the number and type of POIs were found to have a slight impact on the estimation results. On this basis, the authors put forward a method called POI principal component extraction based on supply-demand gap (PPCE-SDG) to select the most significant POIs. The PPCE-SDG involves four steps: k-means clustering (KMC) of blocks based on supply-demand gap; creating a data vector of POIs after counting the POIs in each cluster; extracting the significant POIs through principal component analysis (PCA) of the data vector; importing the extracted POIs to extreme gradient boosting (XGBoost) for OCH supply-demand prediction. Finally, the POIs selected by the PPCE-SDG were proved superior than those collected by other methods in OCH supply-demand estimation, indicating that our model is a desirable tool for significant POIs selection. The research results lay a good basis for the optimization of OCH services.

show abstract

“…Classification and regression trees (CART) is one of the most well-established machine learning techniques, first introduced by Breiman in 1984 [21]. CART is a typical binary tree, its essence is to divide the feature space into two parts and split the scalar attribute and the continuous attribute [22][23][24][25][26]. The CART algorithm consists of the following two steps: (1) Generate decision tree.…”

Section: Classification and Regression Treementioning

confidence: 99%

Research on Predicting Line Loss Rate in Low Voltage Distribution Network Based on Gradient Boosting Decision Tree

Yao

Zhu

et al. 2019

Energies

View full text Add to dashboard Cite

Line loss rate plays an essential role in evaluating the economic operation of power systems. However, in a low voltage (LV) distribution network, calculating line loss rate has become more cumbersome due to poor configuration of the measuring and detecting device, the difficulty in collecting operational data, and the excessive number of components and nodes. Most previous studies mainly focused on the approaches to calculate or predict line loss rate, but rarely involve the evaluation of the prediction results. In this paper, we propose an approach based on a gradient boosting decision tree (GBDT), to predict line loss rate. GBDT inherits the advantages of both statistical models and AI approaches, and can identify the complex and nonlinear relationship while computing the relative importance among variables. An empirical study on a data set in a city demonstrates that our proposed approach performs well in predicting line loss rate, given a large number of unlabeled examples. Experiments and analysis also confirmed the effectiveness of our proposed approach in anomaly detection and practical project management.

show abstract

A Taxi Gap Prediction Method via Double Ensemble Gradient Boosting Decision Tree

Cited by 15 publications

References 10 publications

Using spatio‐temporal deep learning for forecasting demand and supply‐demand gap in ride‐hailing system with anonymised spatial adjacency information

Using spatio‐temporal deep learning for forecasting demand and supply‐demand gap in ride‐hailing system with anonymised spatial adjacency information

Supply-Demand Prediction of DiDi Based on Points of Interests Selection in Extreme Gradient Boosting Algorithm

Research on Predicting Line Loss Rate in Low Voltage Distribution Network Based on Gradient Boosting Decision Tree

Contact Info

Product

Resources

About