2022
Chen, Zhiju; Liu, Kai; Feng, Tao
Examine the Prediction Error of Ride-Hailing Travel Demands with Various Ignored Sparse Demand Effects Journal Article
In: Journal of Advanced Transportation, vol. 2022, 2022.
Abstract | Links | BibTeX | Tags: Ride hailing, Travel demand
@article{Chen2022,
title = {Examine the Prediction Error of Ride-Hailing Travel Demands with Various Ignored Sparse Demand Effects},
author = {Zhiju Chen and Kai Liu and Tao Feng},
url = {https://www.hindawi.com/journals/jat/2022/7690309/},
doi = {https://doi.org/10.1155/2022/7690309},
year = {2022},
date = {2022-04-12},
journal = {Journal of Advanced Transportation},
volume = {2022},
abstract = {The accurate short-term travel demand predictions of ride-hailing orders can promote the optimal dispatching of vehicles in space and time, which is the crucial issue to achieve sustainable development of such dynamic demand-responsive service. The sparse demands are always ignored in the previous models, and the uncertainties in the spatiotemporal distribution of the predictions induced by setting subjective thresholds are rarely explored. This paper attempts to fill this gap and examine the spatiotemporal sparsity effect on ride-hailing travel demand prediction by using Didi Chuxing order data recorded in Chengdu, China. To obtain the spatiotemporal characteristics of the travel demand, three hexagon-based deep learning models (H-CNN-LSTM, H-CNN-GRU, and H-ConvLSTM) are compared by setting various threshold values. The results show that the H-ConvLSTM model has better prediction performance than the others due to its ability to simultaneously capture spatiotemporal features, especially in areas with a high proportion of sparse demands. We found that increasing the minimum demand threshold to delete more sparse data improves the overall prediction accuracy to a certain extent, but the spatiotemporal coverage of the data is also significantly reduced. Results of this study could guide traffic operations in providing better travel services for different regions.},
keywords = {Ride hailing, Travel demand},
pubstate = {published},
tppubtype = {article}
}
The accurate short-term travel demand predictions of ride-hailing orders can promote the optimal dispatching of vehicles in space and time, which is the crucial issue to achieve sustainable development of such dynamic demand-responsive service. The sparse demands are always ignored in the previous models, and the uncertainties in the spatiotemporal distribution of the predictions induced by setting subjective thresholds are rarely explored. This paper attempts to fill this gap and examine the spatiotemporal sparsity effect on ride-hailing travel demand prediction by using Didi Chuxing order data recorded in Chengdu, China. To obtain the spatiotemporal characteristics of the travel demand, three hexagon-based deep learning models (H-CNN-LSTM, H-CNN-GRU, and H-ConvLSTM) are compared by setting various threshold values. The results show that the H-ConvLSTM model has better prediction performance than the others due to its ability to simultaneously capture spatiotemporal features, especially in areas with a high proportion of sparse demands. We found that increasing the minimum demand threshold to delete more sparse data improves the overall prediction accuracy to a certain extent, but the spatiotemporal coverage of the data is also significantly reduced. Results of this study could guide traffic operations in providing better travel services for different regions.
2010
Anggraini, R; Arentze, T A; Timmermans, H J P; Feng, T
Modeling households activity participation decisions in a rule-based system of travel demand Journal Article
In: Journal of the Eastern Asia Society for Transportation Studies, vol. 8, pp. 389-403, 2010, ISSN: 1881-1124.
Abstract | Links | BibTeX | Tags: Activity participation, Activity-based model, Travel demand
@article{Anggraini2010,
title = {Modeling households activity participation decisions in a rule-based system of travel demand},
author = {R Anggraini and T A Arentze and H J P Timmermans and T Feng},
url = {https://www.jstage.jst.go.jp/article/easts/8/0/8_0_389/_article},
doi = {https://doi.org/10.11175/easts.8.389},
issn = {1881-1124},
year = {2010},
date = {2010-01-01},
urldate = {2010-01-01},
journal = {Journal of the Eastern Asia Society for Transportation Studies},
volume = {8},
pages = {389-403},
abstract = {This paper describes an empirical derivation of an activity participation choice model at the household level decisions taking into account the allocated activities and joint activity participation of household heads in discretionary activities. The households that we consider here are two-heads households; each is either a worker or non-worker. Attributes of households, such as, for example the presence of young children, attributes of the work activities and space-time settings are considered as explanatory variables. To deal with this large set of attributes and account for non-linear relationships between the variables, a decision tree induction method – CHAID – is used to derive a decision tree model. We show how the decision tree model can be used as a component in an activity-scheduling model, ALBATROSS, to predict travel demand in an activity-based-micro-simulation system. The model shows a satisfactory performance as indicated by its goodness-of-fit on validation data.},
keywords = {Activity participation, Activity-based model, Travel demand},
pubstate = {published},
tppubtype = {article}
}
This paper describes an empirical derivation of an activity participation choice model at the household level decisions taking into account the allocated activities and joint activity participation of household heads in discretionary activities. The households that we consider here are two-heads households; each is either a worker or non-worker. Attributes of households, such as, for example the presence of young children, attributes of the work activities and space-time settings are considered as explanatory variables. To deal with this large set of attributes and account for non-linear relationships between the variables, a decision tree induction method – CHAID – is used to derive a decision tree model. We show how the decision tree model can be used as a component in an activity-scheduling model, ALBATROSS, to predict travel demand in an activity-based-micro-simulation system. The model shows a satisfactory performance as indicated by its goodness-of-fit on validation data.