Pith. sign in

REVIEW 4 major objections 5 minor 35 references

Next-Generation Travel Demand Modeling with a Generative Framework for Household Activity Coordination

T0 review · 4 major / 5 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 claims that a generative, household-aware model can replace hand-calibrated regional travel demand models while matching their accuracy.

desk verdict A genuinely new learning-based household coordination model with clean activity-level validation, but the headline LA pipeline numbers are weakened by two calibration-to-target steps that need disclosure before they count as predictive. read the letter →

arxiv 2507.08871 v1 pith:VHEXWPNT submitted 2025-07-09 cs.LG cs.AI

Xishun Liao , Haoxuan Ma , Yifan Liu , Yuxiang Wei , Brian Yueshuai He , Chris Stanford , Jiaqi Ma This is my paper · ORCID
classification cs.LGcs.AI
keywords traveldemandmodelingactivity-basedmodelsdeepgenerativehouseholdactivitycoordinationCAMpopulationsynthesisagent-basedsimulationMATSim
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

This paper tries to establish that a fully generative, learning-based pipeline can stand in for a hand-calibrated regional activity-based travel demand model. The central claim is that DeepCAM, a role-aware neural network trained on a national household travel survey, generates daily activity schedules for every household member, including coordinated activities, and that feeding those schedules through location assignment and MATSim traffic simulation reproduces regional travel patterns about as well as the legacy SCAG ABM. The reported numbers are an origin-destination matrix cosine similarity of 0.97 against SCAG, freeway VMT with a JSD of 0.006 and a MAPE of 9.8%, and corridor-level PeMS speed with a JSD of 0.001 and a MAPE of 6.11%. If those numbers hold, a region could build a working travel demand model from census data and a survey without months of manual behavioral calibration, and update it cheaply as conditions change.

What carries the argument

The carrying object is DeepCAM (Deep Coordinated Activity Model), a role-aware multi-person activity generation network. Its defining component is a role-feature matching attention layer: learned role query vectors are softly aligned with household members' attribute embeddings, introducing a diagonal bias into the attention map so each member's predicted activities stay matched to that member's identity, with a residual connection preserving the original embeddings. A transformer encoder then models interdependencies among members, and a decoder conditions each member's next activity on prior activities and person context. The other load-bearing piece is the iterative spatial-refinement update in location assignment, $D_{t+1} = D_t + \eta (F_{\mathrm{target}} - F_{\mathrm{current}})$, which adjusts activity frequencies toward a reference spatial distribution before trips are loaded into the MATSim simulation.

What would settle it

Disable the spatial-refinement update while keeping everything else identical and recompute the origin-destination cosine similarity and freeway VMT JSD; if the OD similarity falls far below 0.97 or the VMT JSD rises far above 0.006, the headline numbers are produced by fitting toward a reference distribution rather than by DeepCAM. A second check is to run the full pipeline on a different region with no local reference distribution or calibration and compare corridor-level speed against observed sensors.

Watch

Extended reading notes

Core claim

The paper's central discovery is that intra-household coordination in daily travel, who travels with whom, when, and for what shared purpose, can be learned from survey data rather than encoded as fixed rules. DeepCAM takes the household head's activity chain, generated by the earlier DeepAM model, plus household and member attributes, and autoregressively produces coordinated chains for the other members through a role-feature matching attention layer that keeps predictions attached to specific individuals and a transformer encoder that lets members influence each other. The loss adds an overconfidence regularizer so the model does not overstate joint participation, and an event table assigns shared event IDs to activities performed together. In the Los Angeles case study the full pipeline, synthetic population, DeepAM head chains, DeepCAM member chains, zonal location assignment with an iterative spatial-refinement update, and calibrated MATSim simulation, produces origin-destination and traffic measures that the paper reports as close to SCAG ABM outputs and to observed PeMS loop-detector data.

Load-bearing premise

The paper never says where the reference distribution used in the spatial-refinement step comes from; the headline origin-destination similarity of 0.97 is only an emergent prediction if that reference is independent of the validation data, and the paper does not establish that independence.

Editorial extensions

If this is right

  • A regional agency could generate a full activity-based travel demand model from census data and a household travel survey, avoiding years of rule and utility calibration, at inference cost on the order of minutes on a single GPU.
  • Household coordination is represented explicitly through shared event IDs, so downstream analysis can ask who participates with whom and which joint activities occur, not just where and when trips happen.
  • Because the modules are decoupled and the activity models are trained on a national survey, the pipeline can be transferred to a new region by swapping in that region's population synthesis and, if needed, local calibration data.
  • The MATSim-based simulation layer turns generated schedules into second-by-second network trajectories, giving the framework the same kind of policy and infrastructure outputs that legacy ABMs currently provide.

Reading between the lines

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

  • Not claimed in the paper: a zero-shot transfer test would separate the learned core from the spatial-refinement fit, by training on the national survey, switching to a second region without any local reference distribution, and checking whether corridor speed JSD stays near 0.001 or degrades.
  • Not claimed in the paper: the event-table design could encode multi-household events as well as household ones, so the coordination mechanism likely extends to friends, coworkers, and community gatherings rather than only family members.
  • Not claimed in the paper: the reported JSD values are computed on marginal distributions, so a stricter validation would compare structural statistics such as trip-chain order, joint-activity membership, and location sequences to test whether entire days are realistic, not just aggregate shares.
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 / 5 minor

Summary. The paper proposes DeepCAM, a transformer-based model for generating household-coordinated daily activity chains, embedded in an end-to-end travel demand pipeline comprising population synthesis, DeepAM for the household head, DeepCAM for other members, zonal location assignment, and MATSim traffic simulation. A Los Angeles case study is used to compare the pipeline with the SCAG ABM and Caltrans PeMS data. The paper's headline quantitative claims are an OD cosine similarity of 0.97 against SCAG, daily network VMT JSD of 0.006 with MAPE 9.8%, and corridor-level traffic speed JSD of 0.001 with MAPE 6.11% against PeMS.

Significance. If the full-pipeline numbers were predictive rather than fitted, the framework would be a substantial practical contribution: it would offer a far cheaper and more scalable alternative to hand-calibrated regional ABMs while reproducing aggregate mobility patterns. The paper's strongest element is the held-out NHTS evaluation of DeepCAM: the aggregate activity-type, timing, duration, participant-count, and role-combination analyses in Figs. 4-6 constitute an independent check of the activity-generation and coordination contribution, and the reported JSD values support the claim that the model captures household coordination. The reported training and inference time figures are also useful engineering evidence. However, the full-pipeline validation currently contains two fit-to-target risks that prevent the headline metrics from being interpreted as predictive accuracy.

major comments (4)
  1. [IV.D.2, V.B.3] The corridor-level validation is circular as presented. Section IV.D.2 states that "Traffic flow dynamics are also calibrated using observed count data from Caltrans PeMS," and Section V.B.3 then uses PeMS as "the ground truth for validation" when reporting corridor speed JSD of 0.001 and volume MAPE of 9.07% (speed MAPE 6.11%). Unless the PeMS detector set and time periods used for calibration are disjoint from the 20 loop detectors and time periods scored in V.B.3, these metrics measure fit to the calibration data rather than predictive accuracy. The paper should specify the calibration and validation detector sets, time ranges, and any preprocessing, and report both in-sample and out-of-sample metrics.
  2. [IV.C.3, V.B.3] The spatial-refinement step in IV.C.3 is another fit-to-target whose provenance is unspecified. The update D_{t+1}=D_t+eta(F_target-F_current) requires a reference spatial distribution F_target, but the paper never says where F_target comes from. If F_target is the SCAG ABM OD matrix or any regional target that also defines the OD benchmark used in V.B.3 (cosine similarity 0.97), then the reported OD agreement is partly enforced by construction rather than an emergent property of the generative model. Please disclose the source of F_target, show that it is independent of the benchmark, or remove the OD comparison from the headline evidence.
  3. [Abstract, V.A] The scale of the demonstration is internally inconsistent. The abstract states that the full-pipeline implementation was conducted "in Los Angeles with a 10 million population," but Section V.A says the framework is applied to "simulate a population of one million residents in the LA region." These numbers cannot both be correct, and the absolute VMT comparison and the scalability/cost claims depend on which is true. Please correct the discrepancy and state clearly how many synthetic agents were simulated in the reported validation.
  4. [V.A, VI] The transferability claim is not supported by any experiment in this paper. The contribution list and abstract describe the framework as "transferable to other regions," and Section V.A appeals to prior DeepAM transfer results, but no experiment in the manuscript demonstrates DeepCAM or the full pipeline on a second region. The claim should be either demonstrated with a cross-region run or explicitly restricted to the DeepAM component.
minor comments (5)
  1. [IV.C.3] The hyperparameters eta and T_max, and the weights alpha and beta, are introduced without values or sensitivity analysis; at least report the values used and indicate whether results are sensitive to them.
  2. [IV.D.1] LASim is mentioned as a framework but not defined; spell out the acronym or cite the corresponding documentation before first use.
  3. [V.B.1] The held-out status of the NHTS evaluation should be stated explicitly in Section V.B.1; the reader should be told that the JSD analyses in Figs. 4-6 are computed on the test split rather than on training data.
  4. [V.B.3] Fig. 7(c) and (d) compare the model with both SCAG and PeMS, but the figure legend and caption do not distinguish which series is PeMS; please label the series directly.
  5. [IV.B.3] The notation \hat{p}_{b,t,p,a} is used before it is defined; move the definition of the indices and variables ahead of the loss formulas.

Circularity Check

2 steps flagged · score 6.0 of 10

OD and PeMS metrics are partly fit-to-target: spatial refinement tunes activity zonal frequencies toward an unspecified F_target before scoring OD against SCAG, and PeMS-calibrated traffic flow is scored against PeMS without a stated holdout; the held-out NHTS evaluation remains an independent check.

  1. fitted input called prediction [Sec. IV.C.3 (Spatial Refinement) and Sec. V.B.3 (Full-Pipeline Simulation Validation)]
    "To match observed spatial distributions D, the activity frequencies are iteratively adjusted: Dt+1 = Dt + η · (Ftarget − Fcurrent), where Fcurrent and Ftarget are the current and reference spatial distributions, and η is a learning rate. ... As shown in Fig. 7(a), the assigned activity locations closely match the result of SCAG ABM across all LA sub-regions, with a cosine similarity of 0.97 between OD matrices."

    The spatial-refinement step is an explicit fitting loop over zonal activity frequencies: D_{t+1} = D_t + eta (F_target - F_current) drives the generated location distribution toward an external reference F_target. The OD validation then reports a cosine similarity of 0.97 between the assigned activity locations and the SCAG ABM OD matrix. The paper never states the source of F_target. If F_target is (or is derived from) the SCAG/regional spatial distribution used in the OD benchmark, the 0.97 cosine similarity is the convergence residual of this update, not an emergent property of DeepCAM. The provenance of F_target must be disclosed before the OD score can be read as a prediction rather than an enforced match.

  2. fitted input called prediction [Sec. IV.D.2 (Simulation Setup) and Sec. V.B.3 (Full-Pipeline Simulation Validation)]
    "Traffic flow dynamics are also calibrated using observed count data from Caltrans PeMS [32]. ... Simulated traffic volume, speed, and flow are compared against real-world observations from the PeMS, which serves as the ground truth."

    The traffic simulation is explicitly calibrated to Caltrans PeMS count data before the full-pipeline validation uses the same PeMS observations as ground truth for corridor speed and volume. The paper does not state that the detectors, time periods, or data splits used for calibration are disjoint from the 20 I-405 loop detectors and hours used for scoring. As a result, the reported corridor JSD of 0.001 for speed and MAPE of 6.11% partly measure agreement with the calibration data, not independent predictive accuracy. This directly affects the abstract's claim that the full pipeline matches observed PeMS corridor conditions and, by extension, the headline that it matches the performance of legacy ABMs.

full rationale

The held-out NHTS evaluation in Figs. 4-6 is a genuine independent check of DeepCAM's activity-chain and household-coordination outputs, and that portion of the paper is not circular. However, two headline full-pipeline metrics are compromised. Sec. IV.C.3's spatial refinement iteratively adjusts activity zonal frequencies toward an unspecified reference distribution F_target, and Sec. V.B.3 then reports an OD cosine similarity of 0.97 against the SCAG ABM OD matrix; if F_target is the SCAG or regional distribution, the score is enforced by the update rather than emergent. Separately, Sec. IV.D.2 calibrates traffic flow dynamics on Caltrans PeMS count data, and Sec. V.B.3 scores the same PeMS data as ground truth without any stated holdout split, so the corridor speed/volume JSD and MAPE partly measure calibration fit. These issues bear directly on the claim that the pipeline matches legacy ABM performance and observed PeMS conditions; they are fixable by disclosing F_target's provenance and using disjoint calibration and validation data, but as written the current numbers cannot distinguish fully predictive from partially fitted performance.

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

The central claim depends on unstated calibration values, the provenance of the spatial refinement target, and transfer of an NHTS-trained model to LA. The largest burdens are the PeMS calibration in Sec IV.D.2 and the F_target fitting in Sec IV.C.3; the activity-level NHTS validation is independent of those.

free parameters (7)
  • lambda_AOR = not reported
    Weight of the Activity Overconfidence Regularization in the DeepCAM loss (Sec IV.B.3); controls the degree of penalty but no value or sensitivity analysis is given.
  • alpha, beta (location assignment weights) = not reported
    Weights in the non-mandatory location assignment objective (Sec IV.C.2) trading distance deviation vs angular deviation; values affect resulting OD patterns.
  • eta (spatial refinement learning rate) = not reported
    Update rate in the spatial refinement loop D_{t+1} = D_t + eta (F_target - F_current) (Sec IV.C.3); directly controls how aggressively locations are fit to the target distribution.
  • T_max (travel time window) = not reported
    Constraint in non-mandatory location assignment (Sec IV.C.2) limiting travel time between consecutive activities.
  • Activity type weights w_a = derived from NHTS
    Used in the overconfidence losses; derived from NHTS rather than listed, and influence the training objective.
  • LASim mode choice utility parameters = calibrated
    Mode choice utility functions 'carefully calibrated' as part of LASim (Sec IV.D.1); values not given.
  • MATSim traffic flow calibration parameters = calibrated to PeMS
    Traffic flow dynamics calibrated using Caltrans PeMS count data (Sec IV.D.2); values not given, and this calibration is the source of the PeMS validation circularity.
assumptions (7)
  • domain assumption Deep learning training on a train/validation/test split of NHTS 2017 is sufficient to learn household activity coordination that transfers to Los Angeles.
    Sec V.A trains DeepAM and DeepCAM on NHTS and then applies them to LA synthetic population; transferability is asserted from prior work [12] rather than re-demonstrated here for coordination.
  • domain assumption Coordinated activities can be identified as household-member activities starting within 15 minutes with the same activity type or an accompanying type.
    Sec IV.B.4 defines coordination this way, following [11], [27]; if this labeling is wrong, the event-table training signal and coordination evaluations inherit the error.
  • domain assumption The household head, chosen by priority on age, employment, license, vehicle access, and gender, provides a valid seed schedule for the household.
    Sec IV.B.1 uses the DeepAM-generated head chain as the seed, citing [18] that household decisions skew toward employed males and primary vehicle users.
  • domain assumption SCAG SimAGENT synthetic population is a valid representation of LA households and is suitable for both benchmark and proposed model.
    Sec V.A adopts the same SCAG population for both models to enable comparison; errors in the synthetic population affect both pipelines equally but still bound the validity of the LA case study.
  • domain assumption MATSim/LASim with calibrated mode choice and traffic flow parameters adequately reproduces LA network dynamics.
    Sec IV.D relies on LASim [31] and calibrates flow dynamics with PeMS counts; the validation inherits this assumption.
  • standard math Standard transformer and gradient-based learning machinery are reliable for sequence generation.
    DeepCAM architecture in Sec IV.B.2 builds on [26]; no formal proof is offered, which is normal for empirical ML.
  • domain assumption Aggregating NHTS activity types from 19 to 15 based on associated locations does not remove the coordination signal being modeled.
    Sec V.A describes the aggregation; the evaluation only covers the 15-type scheme.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Next-Generation Travel Demand Modeling with a Generative Framework for Household Activity Coordination." pith.science (2026). https://pith.science/paper/VHEXWPNT

@misc{pith2026250708871,
  author       = {Pith},
  title        = {Pith review of: Next-Generation Travel Demand Modeling with a Generative Framework for Household Activity Coordination},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VHEXWPNT}},
  note         = {Machine review of arXiv:2507.08871}
}
read the original abstract

Travel demand models are critical tools for planning, policy, and mobility system design. Traditional activity-based models (ABMs), although grounded in behavioral theories, often rely on simplified rules and assumptions, and are costly to develop and difficult to adapt across different regions. This paper presents a learning-based travel demand modeling framework that synthesizes household-coordinated daily activity patterns based on a household's socio-demographic profiles. The whole framework integrates population synthesis, coordinated activity generation, location assignment, and large-scale microscopic traffic simulation into a unified system. It is fully generative, data-driven, scalable, and transferable to other regions. A full-pipeline implementation is conducted in Los Angeles with a 10 million population. Comprehensive validation shows that the model closely replicates real-world mobility patterns and matches the performance of legacy ABMs with significantly reduced modeling cost and greater scalability. With respect to the SCAG ABM benchmark, the origin-destination matrix achieves a cosine similarity of 0.97, and the daily vehicle miles traveled (VMT) in the network yields a 0.006 Jensen-Shannon Divergence (JSD) and a 9.8% mean absolute percentage error (MAPE). When compared to real-world observations from Caltrans PeMS, the evaluation on corridor-level traffic speed and volume reaches a 0.001 JSD and a 6.11% MAPE.

Figures

Figures reproduced from arXiv: 2507.08871 by the authors.

Figure 1
Figure 1. System framework of the proposed next generation [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Network structure of DeepCAM. of the target individual and household members, as well as 13 types of household and zonal characteristics such as vehicle ownership, home ownership, population density, and residential classification. DeepAM generates activity chain autoregressively for each activity Ai,pi until a special end-of-day token is emitted. Built on a transformer-based architecture, DeepAM effec￾tively models… view at source ↗
Figure 3
Figure 3. Traffic simulation for LA using MATSim. 2) Non-Mandatory Activities: Discretionary activities are placed between anchors by minimizing a weighted cost of travel distance and directional deviation, while respecting travel time limits: Znmd = arg minz∈Z n α [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Household activity participation per activity type. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 4
Figure 4. Figure 4: Daily activity patterns: model prediction vs. NHTS. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 6
Figure 6. Figure 6: Most frequent role combinations per activity type. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Validation for activity location assignment and traffic [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 33 canonical work pages

  1. [1]

    Mobility ai agents and networks,

    H. Ma, Y . Liu, Q. Jiang, B. Y . He, X. Liao, and J. Ma, “Mobility ai agents and networks,” IEEE Transactions on Intelligent Vehicles , 2024

    work page 2024

  2. [2]

    An analytical framework to nowcast well-being using mobile phone data,

    L. Pappalardo, M. Vanhoof, L. Gabrielli, Z. Smoreda, D. Pedreschi, and F. Giannotti, “An analytical framework to nowcast well-being using mobile phone data,” International Journal of Data Science and Analytics, vol. 2, pp. 75–92, 2016

    work page 2016

  3. [3]

    Mobility analytics of fans during the 2021 fifa arab cup tm football tournament in qatar,

    J. Abidi and F. Filali, “Mobility analytics of fans during the 2021 fifa arab cup tm football tournament in qatar,” IEEE Open Journal of Intelligent Transportation Systems, 2023

    work page 2021

  4. [4]

    Quantifying the presence of air pollutants over a road network in high spatio-temporal resolution,

    M. Bohm, M. Nanni, and L. Pappalardo, “Quantifying the presence of air pollutants over a road network in high spatio-temporal resolution,” in Climate Change AI, NeurIPS Workshop , 2021

    work page 2021

  5. [5]

    M. E. Ben-Akiva and S. R. Lerman, Discrete choice analysis: theory and application to travel demand . MIT press, 1985, vol. 9

    work page 1985

  6. [6]

    A utility maximizing model of the demand for multi-destination non-work travel,

    J. Horowitz, “A utility maximizing model of the demand for multi-destination non-work travel,” Transportation Research Part B: Methodological, vol. 14, no. 4, pp. 369–386, 1980

    work page 1980

  7. [7]

    Stockholm model system (sims): Application,

    S. Algers, A. Daly, P. Kjellman, and S. Widlert, “Stockholm model system (sims): Application,” in 7th World Conference of Transporta- tion Research. Sydney Australia, 1995, pp. 16–21

    work page 1995

  8. [8]

    Tour based travel demand modeling in the us,

    T. F. Rossi and Y . Shiftan, “Tour based travel demand modeling in the us,” IFAC Proceedings Volumes, vol. 30, no. 8, pp. 381–386, 1997

    work page 1997

Show all 35 references
  1. [9]

    Simulator of activities, greenhouse emissions, networks, and travel (simagent) in southern california: Design, implementation, preliminary findings, and integration plans,

    K. G. Goulias, C. R. Bhat, R. M. Pendyala, Y . Chen, R. Paleti, K. C. Konduri, G. Huang, and H.-H. Hu, “Simulator of activities, greenhouse emissions, networks, and travel (simagent) in southern california: Design, implementation, preliminary findings, and integration plans,” ...

    2011

  2. [10]

    Activity-based disaggregate travel demand model system with activity schedules,

    J. L. Bowman and M. E. Ben-Akiva, “Activity-based disaggregate travel demand model system with activity schedules,” Transportation research part a: policy and practice , vol. 35, no. 1, pp. 1–28, 2001

    2001

  3. [11]

    A household-level activity pattern gen- eration model for the simulator of activities, greenhouse emissions, networks, and travel (simagent) system in southern california,

    C. R. Bhat, K. G. Goulias, R. M. Pendyala, R. Paleti, R. Sidharthan, L. Schmitt, and H.-h. Hu, “A household-level activity pattern gen- eration model for the simulator of activities, greenhouse emissions, networks, and travel (simagent) system in southern california,” in 91st ...

    2012

  4. [12]

    Deep activity model: A generative approach for human mobility pattern synthesis,

    X. Liao, Q. Jiang, B. Y . He, Y . Liu, C. Kuai, and J. Ma, “Deep activity model: A generative approach for human mobility pattern synthesis,” arXiv preprint arXiv:2405.17468 , 2024

    2024 arXiv

  5. [13]

    Sim- mobility: A multi-scale integrated agent-based simulation platform,

    M. Adnan, F. C. Pereira, C. M. L. Azevedo, K. Basak, M. Lovric, S. Raveau, Y . Zhu, J. Ferreira, C. Zegras, and M. Ben-Akiva, “Sim- mobility: A multi-scale integrated agent-based simulation platform,” in 95th Annual Meeting of the Transportation Research Board Forth- coming in...

    2016

  6. [14]

    Activitysim: Activity-based travel de- mand modeling built by and for users (rsg, 2021),

    J. Freedman and D. Hensle, “Activitysim: Activity-based travel de- mand modeling built by and for users (rsg, 2021),” 2021

    2021

  7. [15]

    Modelling activity scheduling behaviour with deep generative machine learning,

    F. Shone and T. Hillel, “Modelling activity scheduling behaviour with deep generative machine learning,” arXiv preprint arXiv:2501.10221 , 2025

    2025

  8. [16]

    Modeling household activity–travel interactions as parallel constrained choices,

    J. P. Gliebe and F. S. Koppelman, “Modeling household activity–travel interactions as parallel constrained choices,” Transportation, vol. 32, pp. 449–471, 2005

    2005

  9. [17]

    A model for joint choice of daily activity pattern types of household members,

    M. Bradley and P. V ovsha, “A model for joint choice of daily activity pattern types of household members,” Transportation, vol. 32, pp. 545–571, 2005

    2005

  10. [18]

    Modeling household discrete choice behavior incorporating heterogeneous group decision- making mechanisms,

    J. Zhang, M. Kuwano, B. Lee, and A. Fujiwara, “Modeling household discrete choice behavior incorporating heterogeneous group decision- making mechanisms,” Transportation Research Part B: Methodologi- cal, vol. 43, no. 2, pp. 230–250, 2009

    2009

  11. [19]

    Joint-activities generation among household members using a latent class model,

    F. Shakeel, M. Adnan, and T. Bellemans, “Joint-activities generation among household members using a latent class model,” Transportation Research Procedia, vol. 62, pp. 557–564, 2022

    2022

  12. [20]

    A need-based model of multi- day, multi-person activity generation,

    T. A. Arentze and H. J. Timmermans, “A need-based model of multi- day, multi-person activity generation,” Transportation Research Part B: Methodological, vol. 43, no. 2, pp. 251–265, 2009

    2009

  13. [21]

    A household-level activity pattern genera- tion model with an application for southern california,

    C. R. Bhat, K. G. Goulias, R. M. Pendyala, R. Paleti, R. Sidharthan, L. Schmitt, and H.-H. Hu, “A household-level activity pattern genera- tion model with an application for southern california,”Transportation, vol. 40, pp. 1063–1086, 2013

    2013

  14. [22]

    Popgen: a virtual human population generator,

    K. McNally, R. Cotton, A. Hogg, and G. Loizou, “Popgen: a virtual human population generator,” Toxicology, vol. 315, pp. 70–85, 2014

    2014

  15. [23]

    PopGen: Synthetic Population Generator,

    Mobility Analytics Research Group, “PopGen: Synthetic Population Generator,” http://www.mobilityanalytics.org/popgen.html, 2016, ac- cessed April 21, 2025

    2016

  16. [24]

    Simagent population synthesis,

    R. M. Pendyala, C. R. Bhat, K. G. Goulias, R. Paleti, K. Konduri, R. Sidharthan, and K. P. Christian, “Simagent population synthesis,” Synthesis, 2012

    2012

  17. [25]

    Simulation based population synthesis,

    B. Farooq, M. Bierlaire, R. Hurtubia, and G. Fl ¨otter¨od, “Simulation based population synthesis,” Transportation Research Part B: Method- ological, vol. 58, pp. 243–263, 2013

    2013

  18. [26]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems , vol. 30, 2017

    2017

  19. [27]

    An integrated spatio-temporal gis toolkit for exploring intra-household interactions,

    H. Kang and D. M. Scott, “An integrated spatio-temporal gis toolkit for exploring intra-household interactions,” Transportation, vol. 35, pp. 253–268, 2008

    2008

  20. [28]

    Horni, K

    A. Horni, K. Nagel, and K. Axhausen, Eds., Multi-Agent Transport Simulation MATSim. London: Ubiquity Press, Aug 2016

    2016

  21. [29]

    Commuting to work in los angeles county,

    Los Angeles Almanac, “Commuting to work in los angeles county,” 2023, accessed: 2025-04-28. [Online]. Available: https://laalmanac.com/employment/em22.php

    2023

  22. [30]

    La metro rail gtfs,

    Los Angeles County Metropolitan Transportation Authority, “La metro rail gtfs,” 2016, accessed: 2025-04-28. [Online]. Available: https://transitfeeds.com/p/la-metro/677?p=176

    2016

  23. [31]

    Multi-agent multimodal transportation simulation for mega-cities: Application of los angeles,

    B. Y . He, Q. Jiang, H. Ma, and J. Ma, “Multi-agent multimodal transportation simulation for mega-cities: Application of los angeles,” 2024

    2024

  24. [32]

    [Online]

    California Department of Transportation, “PeMS,” 2020, retrieved from California Department of Transportation. [Online]. Available: https://pems.dot.ca.gov

    2020

  25. [33]

    National household travel survey,

    F. H. Administration, “National household travel survey,” https://nhts.ornl.gov, 2017

    2017

  26. [34]

    2016 Re- gional Travel Demand Model and Model Validation Report,

    Southern California Association of Governments (SCAG), “2016 Re- gional Travel Demand Model and Model Validation Report,” Southern California Association of Governments, Tech. Rep., 2020

    2016

  27. [35]

    California household travel survey,

    C. D. of Transportation, “California household travel survey,” https://www.nrel.gov/transportation/secure-transportation-data/tsdc- california-travel-survey.html, 2012

    2012

Pith tools

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