Pith. sign in

REVIEW 4 major objections 4 minor 50 references

Robust Vehicle Rebalancing with Deep Uncertainty in Autonomous Mobility-on-Demand Systems

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper introduces DURO, a neural-network uncertainty-interval predictor feeding a robust rebalancing model, claiming it beats deterministic forecasts and matches slower distributionally robust optimization.

desk verdict Sensible idea—learned uncertainty sets for robust rebalancing—but the paper never writes down the actual robust problem it solves: Equation (7) keeps the uncertain demand in the objective and drops the uncertainty constraints. read the letter →

arxiv 2507.04520 v1 pith:4UNLYDQU submitted 2025-07-06 math.OC

Xinling Li , Xiaotong Guo , Qingyi Wang , Gioele Zardini , Jinhua Zhao This is my paper · ORCID
classification math.OC MSC 90B0690C1590C47
keywords Autonomousmobility-on-demandvehiclerebalancingdemanduncertaintyrobustoptimizationdistributionallydeeplearningquantificationgraphneuralnetworks
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a fleet operator can make vehicle-rebalancing decisions under uncertain demand without hand-picking an uncertainty set or paying the computational cost of distributionally robust optimization. DURO trains a graph-convolutional LSTM network to estimate demand distribution parameters, converts them into per-region, per-time prediction intervals, and feeds those intervals directly into a robust optimization model that rebalances idle vehicles and estimates passenger matching. On New York City Manhattan weekday data, the authors report that DURO lowers average passenger waiting time relative to deterministic baselines and is on par with DRO, while solving a full 63-region rebalancing step in under five seconds versus more than five minutes for DRO. If true, real-time robust rebalancing becomes practical, and the same learned-interval-then-robustly-optimize pattern could extend to other decision problems with uncertain inputs. The paper also acknowledges that the tested dataset's regularity may hide DURO's full advantage over baselines.

What carries the argument

The load-bearing object is the per-region, per-time demand interval $(LB_k^i, UB_k^i)$, produced by a probabilistic graph-convolutional LSTM network whose output parametrizes a Poisson demand distribution trained by negative log-likelihood. The interval enters the downstream optimization module through a box constraint, while a polyhedral constraint with a budget parameter $\Gamma$ limits the total demand deviation from the forecast across all regions. This machinery replaces the manually chosen $\rho$ of classical robust optimization with a data-driven interval, preserving linear-programming tractability and avoiding the ambiguity set that makes DRO computationally heavy. The downstream module is a matching-integrated rebalancing model solved on a rolling horizon, which decides rebalancing flows and estimated matching simultaneously while penalizing unsatisfied trips.

What would settle it

Fix DURO's percentile and budget along with the robust baseline's deviation parameter using only data before June 27, then run the simulator on each remaining weekday in June and on several irregular-demand days, such as rainy days or holidays. If DURO's average waiting time and leaving rate are not consistently below the deterministic engines and within DRO's range once parameters are locked, the paper's performance claim would be refuted. A cheaper check is to compute the prediction-interval coverage of the network's intervals on those held-out days: coverage far below the intended 95 percent would show the uncertainty intervals are not faithful.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that demand uncertainty for vehicle rebalancing can be quantified automatically by a neural network, and that the resulting prediction intervals, when plugged into a box-plus-budget robust optimization model, produce rebalancing decisions that beat deterministic forecast-based rebalancing and match distributionally robust optimization at a fraction of its computational cost. In the NYC morning-peak experiments, DURO's average passenger waiting time is lower than both deterministic engines, historical-average demand and neural point prediction, across all simulated parameter combinations. In the head-to-head with DRO on a reduced 30-region setting, the paper reports DRO attaining somewhat lower waiting times and leaving rates, but DURO achieving comparable results once its percentile and budget parameter are chosen, while solving far faster and at full scale. The paper further claims the framework is general, applying to any decision problem with uncertain parameters, not only to autonomous mobility-on-demand rebalancing.

Load-bearing premise

The central claim rests on the assumption that the performance measured on one regular weekday morning in Manhattan, with DURO's uncertainty parameters selected on that same morning, is representative of how the method would perform on other days and less regular demand patterns.

Editorial extensions

If this is right

  • If the central claim holds, fleet operators no longer need to hand-craft robust optimization uncertainty sets: the neural network supplies per-region intervals, and only a percentile and a budget need to be set, both with clearer operational meaning than a raw deviation parameter.
  • A fleet running DURO on the tested 63-region Manhattan case can make a rebalancing decision in under five seconds per interval, whereas DRO required more than five minutes even after restricting rebalancing to 30 high-demand regions.
  • DURO's average passenger waiting time is reported lower than both deterministic engines, historical-average demand and neural point forecast, across all simulated parameter combinations in the morning-peak test.
  • In the direct comparison with DRO on the reduced 30-region setting, DURO is within a small performance gap once its percentile and budget parameter are chosen, and the paper frames this gap as an acceptable trade for the large computational saving.
  • The framework is presented as general-purpose rather than AMoD-specific: any decision problem with uncertain parameters could use the same architecture of learned uncertainty intervals upstream of a robust optimization model.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, the same learned-interval-plus-robust-optimization pattern could be transplanted to other real-time decisions with uncertain inputs, such as ride-hailing pricing, charging-station dispatch, or warehouse task assignment, provided the downstream problem is a linear or convex program that accepts box-plus-budget uncertainty.
  • Beyond the paper, the performance comparison would be more decisive on days when the historical average is a poor forecast; the paper itself notes its weekdays-only, regular-demand data may hide DURO's advantage, so testing on rainy days, holidays, or weekends is a natural next experiment.
  • Beyond the paper, one could decouple the network's contribution from the distributional assumption by replacing the Poisson likelihood with Bayesian quantile regression, a direction the authors suggest, and comparing interval sharpness and final rebalancing metrics; this would show how much of the gain comes from learning versus from the chosen distribution.
  • Beyond the paper, a validation-style experiment that fixes the percentile and budget on a validation set and then evaluates on unseen test days would quantify how much tuning effort actually remains in DURO, since the reported results use the best-performing parameters selected on the test morning.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes DURO, a vehicle-rebalancing framework for autonomous mobility-on-demand systems in which a graph-LSTM neural network predicts distributional parameters of demand, the predicted distribution is turned into a box-plus-budget uncertainty set, and a robust optimization model is solved in a model-predictive-control loop. The authors benchmark DURO against deterministic optimization with historical-average and neural-network point forecasts, against classical robust optimization with a manually specified uncertainty set, and against a moment-based distributionally robust optimization model. Experiments on Manhattan taxi data from June 2019 report that DURO reduces passenger waiting time and leaving rate relative to deterministic baselines, is competitive with RO and DRO, and is substantially faster than DRO.

Significance. If the claimed results held, the paper would make a useful practical contribution: it couples a data-driven uncertainty-quantification module with robust optimization, avoids manual uncertainty-set tuning in RO, and avoids the computational cost of DRO. The idea of feeding neural-network-derived prediction intervals into a robust rebalancing model is timely and relevant. However, the supporting evidence is currently incomplete: the robust counterpart of the optimization model is not derived, the numerical comparison is based on a single 2-hour test window with parameters selected on that same window, and the DRO comparison is run under a different, reduced setting. These issues make the central claim 'on par with DRO, but with superior computational efficiency' not yet established. The paper does not provide code or machine-checked proofs, but the empirical workflow is clearly described and the authors honestly acknowledge in the conclusion that the regularity of their dataset obscures DURO's potential.

major comments (4)
  1. [Section III.C, Eq. (7)] Equation (7) is not a robust counterpart of the model with uncertainty set (6a)-(6b): it still contains the uncertain demand r_i directly in the objective, and no worst-case maximization, epigraph reformulation, or dual transformation is provided. Since the objective coefficient on r_i is positive through the unsatisfied-trip term gamma * T_i, the adversarial demand is not trivially the upper bound when the budget constraint (6b) is active. The actual optimization problem solved in the experiments is therefore undefined, and the numerical results in Table II and Figures 10-14 cannot be reproduced or audited. This is a load-bearing gap that must be fixed before the method can be evaluated.
  2. [Section IV.D, Figs. 10-13] The paper reports the best-performing parameter combination (PI=75%, Gamma=9 for DURO; the best rho for RO) chosen by evaluating on the same single test morning (June 27, 7:00-9:00) that is used for all comparisons. This is selection on the test set, and with only one 2-hour window there are no repeated trials or error bars to quantify variability. The claim that DURO consistently outperforms DOHV and DONN, and is competitive with RO, is therefore not supported by the presented evidence. The authors' own conclusion that the dataset's regularity 'obscure[s] the full potential of DURO' reinforces the need for evaluation on multiple days or repeated demand scenarios before drawing general conclusions.
  3. [Section IV.D.6, Table II, Fig. 14] The DRO comparison is not apples-to-apples. In Table II the DRO* entry rebalances only in 'a few regions' because of computational complexity, which does not represent the same control problem as the other engines. In the later DRO comparison the number of rebalancing regions is reduced from 63 to 30, the ambiguity set is augmented with neighboring 5-minute data to achieve feasibility, and DURO is run with only the 75% and 95% intervals because the 50% interval is infeasible. These changes mean the DRO results in Fig. 14 are obtained in a different experimental setting from the other benchmark results, and the conclusion that DURO is 'on par with DRO' is not justified by the reported evidence.
  4. [Section III.C, Constraints (3b) and (4a)-(4d)] The reformulation issue also affects the constraints: constraint (3b), y_sum <= r_i, contains the uncertain r_i, and the transition constraints (4a)-(4d) couple rebalancing and matching decisions across look-ahead intervals. Even if one accepted Eq. (7) as the objective, the robust version of the full feasible set must be specified, for example by enforcing the constraints for all r in the uncertainty set or by moving to a two-stage formulation. The paper does not state which robust semantics (here-and-now, adjustable, or constraint-wise) are being used, so the reported optimization results are unverifiable.
minor comments (4)
  1. [Global] There are multiple typos and formatting artifacts, including 'constrains' for 'constraints', 'X-ais' for 'X-axis', 'Possion' for 'Poisson', and the unresolved LaTeX macro 'glsabk:dro' in Section IV.D.6. These should be corrected before publication.
  2. [Section III.C, objective (5)] The objective (5) is written as a formula followed by a sentence that explains beta and gamma, but the equation itself is not numbered consistently with its reference in the text; the same applies to the display for (7). Please number all displayed equations and refer to them consistently.
  3. [Section IV.D.5] The sentence 'this result serves as the baseline for later comparison' appears to refer to Fig. 10, but the baseline terminology is ambiguous because later comparisons report percentage reductions relative to DOHV and DONN, not to the DURO parameter sweep. Clarify what the baseline is in each comparison.
  4. [Section V] The conclusion describes the uncertainty set of RO as 'momentum-based', while Section III.A uses the same phrase for the framework in [41]. This is likely a typo for 'moment-based' or 'momentum-based' in a different sense. Please align the terminology with the body of the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: DURO's uncertainty intervals are trained on historical demand and evaluated on future realized demand; the self-cited base model [12] is restated and is not the load-bearing evidence for the central claim.

full rationale

The derivation chain is not circular. The demand uncertainty intervals are produced by a graph-LSTM network trained on June 1-24 demand via negative log-likelihood, with validation on June 25-26 and testing on June 27; these intervals enter the robust rebalancing model as inputs, and the reported waiting times and leaving rates are generated by a simulator that uses realized demand only for passenger generation. No final-performance parameter is fitted inside the optimization, and the network is not trained on waiting time. The base matching-integrated rebalancing model and the static transition matrices are attributed to the authors' prior work [12], but the equations are restated in the manuscript, and [12] is an externally published reference rather than an unverified uniqueness theorem. The self-citations [12], [24], and [41] are background or component citations and do not carry the central claim that learned uncertainty sets improve robust rebalancing. The main caveats are methodological rather than circular: the best-performing PI and Gamma for DURO and rho for RO are selected on the same single test morning and the paper reports 'the result of the best-performing one' in Section IV.D, which is in-sample selection bias, not circularity. In addition, Equation (7) is presented as a reframed robust model but still contains the uncertain variable r_i with a positive objective coefficient and no worst-case or dual reformulation, so the actual solved optimization is underspecified; this is an auditability and correctness concern, not an equivalence between output and input by construction. The conclusion itself concedes that the dataset's regularity 'obscure[s] the full potential of DURO' (Section V), further confirming that the reported comparisons are limited by benchmark design rather than by circular derivation. Under the stated circularity definition, this is a normal non-circular finding.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central claim rests on a standard robust-optimization reformulation that is asserted but not derived, on distributional assumptions for the uncertainty intervals, on static transition matrices inherited from prior work, and on a benchmark protocol that selects best parameters on the test day. No new physical entities are introduced.

free parameters (7)
  • Gamma (budget in polyhedral uncertainty constraint) = 9 (best for PI=75% on test day)
    Swept over values; best-performing choice selected on the June 27 test set, Section IV.D.2 and Fig 10.
  • PI (prediction interval percentile) = 75% (best)
    Swept over 50%, 75%, 95%; authors note the optimal PI is problem-specific, chosen on the test set.
  • Beta (pickup distance weight) = 1
    Hand-set in Table I; no sensitivity analysis reported.
  • Gamma penalty for unsatisfied requests = 100
    Hand-set in Table I; no sensitivity analysis reported.
  • Rho (RO uncertainty parameter) = varied; best case reported
    Manual RO benchmark parameter; authors use its tuning burden as motivation for DURO.
  • Neural network training hyperparameters = not reported
    Architecture (two GCN layers plus LSTM) is described, but hidden sizes, learning rate, epochs, and seeds are unstated; these choices affect interval quality.
  • Static transition matrices P and Q = estimated from historical data
    Used in constraints (4c) and (4d); approximation from historical data with no uncertainty or estimation variance reported.
assumptions (6)
  • standard math The box-plus-budget robust optimization model can be reframed into the stated tractable form without loss.
    Section III.C states 'the robust optimization model can be reframed into the following format' but gives no dualization or derivation; the displayed Eq. (7) still contains the uncertain variable r_i^k.
  • domain assumption Demand follows the chosen probabilistic distribution, and NLL-trained intervals are calibrated.
    Section III.B and IV.D.1; Poisson is selected empirically, and if the distributional assumption is wrong the uncertainty intervals are miscalibrated.
  • domain assumption Static transition matrices P and Q adequately model occupied vehicle movement.
    Section III.C: 'We approximate them with static matrices estimated from historical data. Interested readers can refer to [12]'.
  • domain assumption Passenger behavior is captured by a fixed 300-second maximum waiting time computed on 30-second matching steps.
    Section IV.B describes the simulator; this simplification affects all engines equally but may not reflect real abandonment behavior.
  • ad hoc to paper Adjacent 5-minute demand data can be pooled to make the DRO ambiguity set feasible.
    Section IV.D.6 augments historical demand with data from the previous and next 5 minutes, assuming minimal demand change; this is applied only to the DRO baseline.
  • ad hoc to paper The best-performing parameter combination on the test day is a fair representation of each engine's capability.
    Section IV.D: 'For engines with various parameter combination cases, we show the result of the best-performing one.' This is an evaluation protocol assumption, not a mathematical axiom.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Vehicle Rebalancing with Deep Uncertainty in Autonomous Mobility-on-Demand Systems." pith.science (2026). https://pith.science/paper/4UNLYDQU

@misc{pith2026250704520,
  author       = {Pith},
  title        = {Pith review of: Robust Vehicle Rebalancing with Deep Uncertainty in Autonomous Mobility-on-Demand Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4UNLYDQU}},
  note         = {Machine review of arXiv:2507.04520}
}
read the original abstract

Autonomous Mobility-on-Demand (AMoD) services offer an opportunity for improving passenger service while reducing pollution and energy consumption through effective vehicle coordination. A primary challenge in the autonomous fleets coordination is to tackle the inherent issue of supply-demand imbalance. A key strategy in resolving this is vehicle rebalancing, strategically directing idle vehicles to areas with anticipated future demand. Traditional research focuses on deterministic optimization using specific demand forecasts, but the unpredictable nature of demand calls for methods that can manage this uncertainty. This paper introduces the Deep Uncertainty Robust Optimization (DURO), a framework specifically designed for vehicle rebalancing in AMoD systems amidst uncertain demand based on neural networks for robust optimization. DURO forecasts demand uncertainty intervals using a deep neural network, which are then integrated into a robust optimization model. We assess DURO against various established models, including deterministic optimization with refined demand forecasts and Distributionally Robust Optimization (DRO). Based on real-world data from New York City (NYC), our findings show that DURO surpasses traditional deterministic models in accuracy and is on par with DRO, but with superior computational efficiency. The DURO framework is a promising approach for vehicle rebalancing in AMoD systems that is proven to be effective in managing demand uncertainty, competitive in performance, and more computationally efficient than other optimization models.

Figures

Figures reproduced from arXiv: 2507.04520 by the authors.

Figure 4
Figure 4. DURO workflow. A neural network is leveraged to [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 2
Figure 2. Optimization workflow proposed by [41]. The mean [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. DRO workflow. The historical data is used to build [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (7 more)
Figure 7
Figure 7. Figure 7: Example of rolling optimization. The green interval [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]
Figure 8
Figure 8. Figure 8: Demand pattern of NYC Manhattan. The demand is [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: Simulation workflow in the experiments. At each [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Performance of DURO based on Poisson distribution. [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 11
Figure 11. Figure 11: Comparison between DURO and DOHV. DURO achieves a lower average waiting time and passenger leaving rate across all instances compared to DOHV. 3) Comparison with DOHV The historical average demand is a reasonable approxima￾tion of the true demand in cases with regular…
Figure 14
Figure 14. Figure 14: Comparison between DURO and DRO. DRO outper [PITH_FULL_IMAGE:figures/full_fig_p011_14.png]
Figure 13
Figure 13. Figure 13: Comparison between DURO and RO. In most of the [PITH_FULL_IMAGE:figures/full_fig_p011_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 48 canonical work pages

  1. [12]

    Robust matching-integrated vehicle rebalancing in ride-hailing system with uncertain demand,

    X. Guo, N. S. Caros, and J. Zhao, “Robust matching-integrated vehicle rebalancing in ride-hailing system with uncertain demand,” Transporta- tion Research Part B: Methodological , vol. 150, pp. 161–189, 2021

    work page 2021

  2. [1]

    How machine learning informs ride-hailing services: A survey,

    Y . Liu, R. Jia, J. Ye, and X. Qu, “How machine learning informs ride-hailing services: A survey,” Communications in Transportation Research, vol. 2, p. 100075, 2022

    work page 2022

  3. [2]

    Reinforcement learning for rideshar- ing: An extended survey,

    Z. T. Qin, H. Zhu, and J. Ye, “Reinforcement learning for rideshar- ing: An extended survey,” Transportation Research Part C: Emerging Technologies, vol. 144, p. 103852, 2022, review

    work page 2022

  4. [3]

    Robo-taxi Fleet Coordination at Scale via Reinforcement Learning

    L. Tresca, C. Schmidt, J. Harrison, F. Rodrigues, G. Zardini, D. Gam- melli, and M. Pavone, “Robo-taxi fleet coordination at scale via rein- forcement learning,” arXiv preprint arXiv:2504.06125 , 2025

    work page Pith review arXiv 2025

  5. [4]

    Mobility-on-demand pricing versus private vehicle tco: how cost structures hinder the dethroning of the car,

    T. Kuhnimhof and C. Eisenmann, “Mobility-on-demand pricing versus private vehicle tco: how cost structures hinder the dethroning of the car,” Transportation, vol. 50, no. 2, pp. 707–731, 2023

    work page 2023

  6. [5]

    Analysis and control of autonomous mobility-on-demand systems,

    G. Zardini, N. Lanzetti, M. Pavone, and E. Frazzoli, “Analysis and control of autonomous mobility-on-demand systems,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 5, no. 1, pp. 633–658, 2022

    work page 2022

  7. [6]

    Deep multi-view spatial-temporal network for taxi demand prediction,

    H. Yao, F. Wu, J. Ke, X. Tang, Y . Jia, S. Lu, P. Gong, J. Ye, and Z. Li, “Deep multi-view spatial-temporal network for taxi demand prediction,” in Proceedings of the AAAI conference on artificial intelligence , vol. 32, no. 1, 2018

    work page 2018

  8. [7]

    Traffic demand prediction based on dynamic transition convolutional neural network,

    B. Du, X. Hu, L. Sun, J. Liu, Y . Qiao, and W. Lv, “Traffic demand prediction based on dynamic transition convolutional neural network,” IEEE Transactions on Intelligent Transportation Systems, vol. 22, no. 2, pp. 1237–1247, 2020

    work page 2020

Show all 50 references
  1. [8]

    A dynamical spatial-temporal graph neural network for traffic demand prediction,

    F. Huang, P. Yi, J. Wang, M. Li, J. Peng, and X. Xiong, “A dynamical spatial-temporal graph neural network for traffic demand prediction,” Information Sciences, vol. 594, pp. 286–304, 2022

    2022

  2. [9]

    Approximate dynamic programming for planning a ride-hailing system using autonomous fleets of electric vehicles,

    L. Al-Kanj, J. Nascimento, and W. B. Powell, “Approximate dynamic programming for planning a ride-hailing system using autonomous fleets of electric vehicles,” European Journal of Operational Research , vol. 284, no. 3, pp. 1088–1106, 2020

    2020

  3. [10]

    Dynamic modeling and control of taxi services in large-scale urban networks: A macroscopic approach,

    M. Ramezani and M. Nourinejad, “Dynamic modeling and control of taxi services in large-scale urban networks: A macroscopic approach,” Transportation Research Part C: Emerging Technologies , vol. 94, pp. 203–219, 2018, iSTTT22

    2018

  4. [11]

    Data-driven distributionally robust electric vehicle balancing for mobility-on-demand systems under demand and supply uncertainties,

    S. He, L. Pepin, G. Wang, D. Zhang, and F. Miao, “Data-driven distributionally robust electric vehicle balancing for mobility-on-demand systems under demand and supply uncertainties,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2020, pp...

    2020

  5. [13]

    Ben-Tal, L

    A. Ben-Tal, L. El Ghaoui, and A. Nemirovski, Robust optimization . Princeton university press, 2009, vol. 28

    2009

  6. [14]

    Frameworks and results in distribution- ally robust optimization,

    H. Rahimian and S. Mehrotra, “Frameworks and results in distribution- ally robust optimization,” Open Journal of Mathematical Optimization , vol. 3, p. 1–85, Jul. 2022

    2022

  7. [15]

    Gradient-based learning applied to document recognition,

    Y . LeCun, L. Bottou, Y . Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE , vol. 86, no. 11, pp. 2278–2324, 1998. 13

    1998

  8. [16]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,” vol. 9, no. 8, p. 1735–1780, nov 1997

    1997

  9. [17]

    Semi-supervised classification with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” arXiv preprint arXiv:1609.02907 , 2016

    2016 arXiv

  10. [18]

    Spatio-temporal graph convolutional net- works: A deep learning framework for traffic forecasting,

    B. Yu, H. Yin, and Z. Zhu, “Spatio-temporal graph convolutional net- works: A deep learning framework for traffic forecasting,” in Proceed- ings of the Twenty-Seventh International Joint Conference on Artificial Intelligence. International Joint Conferences on Artificial Intel...

    2018

  11. [19]

    Gman: A graph multi-attention network for traffic prediction,

    C. Zheng, X. Fan, C. Wang, and J. Qi, “Gman: A graph multi-attention network for traffic prediction,” 2019

    2019

  12. [20]

    A review of demand forecasting models and methodological develop- ments within tourism and passenger transportation industry,

    I. Ghalehkhondabi, E. Ardjmand, W. A. Young, and G. R. Weckman, “A review of demand forecasting models and methodological develop- ments within tourism and passenger transportation industry,” Journal of Tourism Futures, vol. 5, no. 1, pp. 75–93, 2019

    2019

  13. [21]

    Spatiotemporal traffic forecasting: review and proposed directions,

    A. Ermagun and D. Levinson, “Spatiotemporal traffic forecasting: review and proposed directions,” Transport Reviews, vol. 38, no. 6, pp. 786– 814, 2018

    2018

  14. [22]

    The propagation of uncertainty through travel demand models: an exploratory analysis,

    Y . Zhao and K. M. Kockelman, “The propagation of uncertainty through travel demand models: an exploratory analysis,” The Annals of regional science, vol. 36, pp. 145–163, 2002

    2002

  15. [23]

    Uncertainty quantification of sparse travel demand prediction with spatial-temporal graph neural networks,

    D. Zhuang, S. Wang, H. Koutsopoulos, and J. Zhao, “Uncertainty quantification of sparse travel demand prediction with spatial-temporal graph neural networks,” in Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2022, pp. 4639– 4647

    2022

  16. [24]

    Uncer- tainty quantification of spatiotemporal travel demand with probabilistic graph neural networks,

    Q. Wang, S. Wang, D. Zhuang, H. Koutsopoulos, and J. Zhao, “Uncer- tainty quantification of spatiotemporal travel demand with probabilistic graph neural networks,” 2023

    2023

  17. [25]

    Beyond expectation: Deep joint mean and quantile regression for spatiotemporal problems,

    F. Rodrigues and F. C. Pereira, “Beyond expectation: Deep joint mean and quantile regression for spatiotemporal problems,” IEEE transactions on neural networks and learning systems, vol. 31, no. 12, pp. 5377–5389, 2020

    2020

  18. [26]

    Temporal latent auto-encoder: A method for probabilistic multivariate time series forecasting,

    N. Nguyen and B. Quanz, “Temporal latent auto-encoder: A method for probabilistic multivariate time series forecasting,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 35, no. 10, 2021, pp. 9117–9125

    2021

  19. [27]

    J. R. Birge and F. Louveaux, Introduction to stochastic programming . Springer Science & Business Media, 2011

    2011

  20. [28]

    Facility location: A robust op- timization approach,

    O. Baron, J. Milner, and H. Naseraldin, “Facility location: A robust op- timization approach,” Production and Operations Management, vol. 20, no. 5, pp. 772–785, 2011

    2011

  21. [29]

    Robust strategies for facility location under uncertainty,

    N. G ¨ulpınar, D. Pachamanova, and E. C ¸ anako˘glu, “Robust strategies for facility location under uncertainty,” European Journal of Operational Research, vol. 225, no. 1, pp. 21–35, 2013

    2013

  22. [30]

    Computing robust basestock levels,

    D. Bienstock and N. ¨Ozbay, “Computing robust basestock levels,” Discrete Optimization, vol. 5, no. 2, pp. 389–414, 2008

    2008

  23. [31]

    Robust approximation to multiperiod inventory management,

    C.-T. See and M. Sim, “Robust approximation to multiperiod inventory management,” Operations research, vol. 58, no. 3, pp. 583–594, 2010

    2010

  24. [32]

    Distributionally robust optimization under mo- ment uncertainty with application to data-driven problems,

    E. Delage and Y . Ye, “Distributionally robust optimization under mo- ment uncertainty with application to data-driven problems,” Operations research, vol. 58, no. 3, pp. 595–612, 2010

    2010

  25. [33]

    Adaptive distributionally robust optimization,

    D. Bertsimas, M. Sim, and M. Zhang, “Adaptive distributionally robust optimization,” Management Science, vol. 65, no. 2, pp. 604–618, 2019

    2019

  26. [34]

    Ridesourcing systems: A framework and review,

    H. Wang and H. Yang, “Ridesourcing systems: A framework and review,” Transportation Research Part B: Methodological, vol. 129, pp. 122–155, 2019

    2019

  27. [35]

    Shared-vehicle mobility-on-demand systems: a fleet operator’s guide to rebalancing empty vehicles,

    K. Spieser, S. Samaranayake, W. Gruel, and E. Frazzoli, “Shared-vehicle mobility-on-demand systems: a fleet operator’s guide to rebalancing empty vehicles,” in Transportation Research Board 95th Annual Meet- ing, no. 16-5987. Transportation Research Board, 2016

    2016

  28. [36]

    Vehicle rebalancing for mobility-on-demand systems with ride-sharing,

    A. Wallar, M. Van Der Zee, J. Alonso-Mora, and D. Rus, “Vehicle rebalancing for mobility-on-demand systems with ride-sharing,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2018, pp. 4539–4546

    2018

  29. [37]

    Robotic load balancing for mobility-on-demand systems,

    M. Pavone, S. L. Smith, E. Frazzoli, and D. Rus, “Robotic load balancing for mobility-on-demand systems,” The International Journal of Robotics Research, vol. 31, no. 7, pp. 839–854, 2012

    2012

  30. [38]

    Control of robotic mobility-on-demand systems: a queueing-theoretical perspective,

    R. Zhang and M. Pavone, “Control of robotic mobility-on-demand systems: a queueing-theoretical perspective,” 2014

    2014

  31. [39]

    Data-driven distributionally robust vehicle balancing using dynamic region partitions,

    F. Miao, S. Han, A. M. Hendawi, M. E. Khalefa, J. A. Stankovic, and G. J. Pappas, “Data-driven distributionally robust vehicle balancing using dynamic region partitions,” in Proceedings of the 8th International Conference on Cyber-Physical Systems, ser. ICCPS ’17. New York, NY...

    2017

  32. [40]

    Rebalancing shared mobility-on- demand systems: A reinforcement learning approach,

    J. Wen, J. Zhao, and P. Jaillet, “Rebalancing shared mobility-on- demand systems: A reinforcement learning approach,” in 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC), 2017, pp. 220–225

    2017

  33. [41]

    Data-driven vehicle rebalancing with predictive prescriptions in the ride-hailing system,

    X. Guo, Q. Wang, and J. Zhao, “Data-driven vehicle rebalancing with predictive prescriptions in the ride-hailing system,” IEEE Open Journal of Intelligent Transportation Systems , vol. 3, pp. 251–266, 2022

    2022

  34. [42]

    Risk-aware operation modeling for ride-hailing fleet in order grabbing mode: A distributional reinforce- ment learning approach,

    Y . Sun, Z. Ding, Z. Hu, and W.-J. Lee, “Risk-aware operation modeling for ride-hailing fleet in order grabbing mode: A distributional reinforce- ment learning approach,” IEEE Transactions on Smart Grid , pp. 1–1, 2023

    2023

  35. [43]

    Data-driven robust taxi dispatch under demand uncertainties,

    F. Miao, S. Han, S. Lin, Q. Wang, J. A. Stankovic, A. Hendawi, D. Zhang, T. He, and G. J. Pappas, “Data-driven robust taxi dispatch under demand uncertainties,” IEEE Transactions on Control Systems Technology, vol. 27, no. 1, pp. 175–191, 2017

    2017

  36. [44]

    Data-driven distributionally robust vehicle balancing using dynamic region partitions,

    F. Miao, S. Han, A. M. Hendawi, M. E. Khalefa, J. A. Stankovic, and G. J. Pappas, “Data-driven distributionally robust vehicle balancing using dynamic region partitions,” in Proceedings of the 8th International Conference on Cyber-Physical Systems , Conference Proceedings, pp. 261–271

  37. [45]

    Time series forecasting using a hybrid arima and neural network model,

    G. Zhang, “Time series forecasting using a hybrid arima and neural network model,” Neurocomputing, vol. 50, pp. 159–175, 2003

    2003

  38. [46]

    Data-driven robust optimization for contextual vehicle rebalancing in on-demand ride services under demand uncertainty,

    Z. Guo, B. Yu, W. Shan, and B. Yao, “Data-driven robust optimization for contextual vehicle rebalancing in on-demand ride services under demand uncertainty,” Transportation Research Part C: Emerging Tech- nologies, vol. 154, p. 104244, 2023

    2023

  39. [47]

    Graph neural network for traffic forecasting: A survey,

    W. Jiang and J. Luo, “Graph neural network for traffic forecasting: A survey,” Expert Systems with Applications , vol. 207, p. 117921, 2022

    2022

  40. [48]

    Lower upper bound estimation method for construction of neural network-based prediction intervals,

    A. Khosravi, S. Nahavandi, D. Creighton, and A. F. Atiya, “Lower upper bound estimation method for construction of neural network-based prediction intervals,” IEEE Transactions on Neural Networks , vol. 22, no. 3, pp. 337–346, 2011

    2011

  41. [49]

    Distributionally robust convex optimization,

    W. Wiesemann, D. Kuhn, and M. Sim, “Distributionally robust convex optimization,” Operations research, vol. 62, no. 6, pp. 1358–1376, 2014

    2014

  42. [50]

    Robust stochastic optimization made easy with rsome,

    Z. Chen, M. Sim, and P. Xiong, “Robust stochastic optimization made easy with rsome,” Management Science, vol. 66, no. 8, pp. 3329–3339, 2020

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.