pith. sign in

arxiv: 2606.01640 · v1 · pith:KVIWWAFEnew · submitted 2026-06-01 · 💻 cs.AI · cs.CL

MobEvolve: An Agentic Self-Evolving Heuristic System for Interpretable Human Mobility Generation

Junlin He , Yihong Tang , Tong Nie , Ao Qu , Yuebing Liang , Hamzeh Alizadeh , Bang Liu , Wei Ma
show 1 more author
Lijun Sun
This is my paper

Pith reviewed 2026-06-28 14:45 UTC · model grok-4.3

classification 💻 cs.AI cs.CL
keywords human mobility generationself-evolving heuristicsLLM agenttrajectory synthesisinterpretabilitybehavioral modelingagentic frameworkpopulation distribution
0
0 comments X

The pith

An LLM agent iteratively evolves a heuristic system to generate human mobility trajectories with higher fidelity and alignment than deep models.

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

MobEvolve starts with behavior-inspired heuristics for synthesizing individual trip chains and deploys an LLM agent to analyze validation set failures. The agent proposes logic updates and builds a memory of changes to improve the system over iterations. This setup targets the gap where deep models sacrifice interpretability and traditional heuristics lack adaptability. If the approach works, mobility generation can achieve high accuracy without black-box models or manual rule tuning. Sympathetic readers would value this for applications like city planning that require both precise simulations and transparent logic.

Core claim

MobEvolve initializes a behavior-inspired heuristic system and employs an LLM agent to iteratively evolve its internal logic. By diagnosing empirical misalignments and failure cases on a validation set, the agent proposes targeted updates and accumulates evolution memory for cumulative self-improvement. On Singapore and Montreal benchmarks, it outperforms state-of-the-art deep generative and LLM-based methods in individual trajectory fidelity, population-level distribution alignment, and behavioral plausibility, while preserving interpretability and high inference efficiency.

What carries the argument

The LLM agent that diagnoses misalignments on validation data and proposes targeted updates to the heuristic logic while accumulating evolution memory.

If this is right

  • Individual trajectories match real data more closely than prior methods.
  • Generated populations align better with observed aggregate statistics.
  • Generated trips exhibit higher behavioral plausibility under domain checks.
  • The generation process remains inspectable through its explicit rule structure.
  • Inference speed stays high relative to deep generative alternatives.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The self-evolution approach could extend to other heuristic-based simulation domains like traffic flow or epidemic spread.
  • Accumulated evolution memory might support quick adaptation when deploying the system to new cities or regions.
  • If agent reliability improves, the method could reduce manual effort needed to maintain and update mobility models.

Load-bearing premise

The LLM agent can reliably identify empirical misalignments on the validation set and propose targeted, non-regressive updates to the heuristic logic without introducing new biases or hallucinations.

What would settle it

Observing no gain or a decline in fidelity and alignment metrics after multiple evolution iterations on the Montreal benchmark would indicate the central claim does not hold.

Figures

Figures reproduced from arXiv: 2606.01640 by Ao Qu, Bang Liu, Hamzeh Alizadeh, Junlin He, Lijun Sun, Tong Nie, Wei Ma, Yihong Tang, Yuebing Liang.

Figure 1
Figure 1. Figure 1: Overview of MobEvolve. Top: feature￾conditioned synthesis of typical-day trip chains. Bottom: MobEvolve brings together interpretability, behavioral plausibility, distributional alignment, and efficiency. trajectories (Wu et al., 2021; Zhang and Dai, 2018; Long et al., 2025), our task focuses on a feature￾conditioned synthesis problem, often operating un￾der a regime where only limited paired observa￾tions… view at source ↗
Figure 2
Figure 2. Figure 2: Overview of MobEvolve. The upper panel illustrates the interpretable mobility heuristic system, which [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Self-evolution trajectory on the Singapore [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Implementation organization of MobEvolve. The frozen components define the task and evaluator, [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
read the original abstract

Human mobility generation aims to synthesize realistic trip chains for target populations based on individual features. Existing paradigms, including deep generative models, LLM-based methods, and traditional heuristics, struggle to satisfy the complex demands of this task while simultaneously maintaining interpretability, behavioral plausibility, population-level distributional alignment, and inference efficiency. To bridge this gap, we introduce MobEvolve, the first agentic self-evolving heuristic framework for human mobility generation. MobEvolve initializes a behavior-inspired heuristic system and employs an LLM agent to iteratively evolve its internal logic. By diagnosing empirical misalignments and failure cases on a validation set, the agent proposes targeted updates and accumulates evolution memory for cumulative self-improvement. Extensive evaluations on the Singapore and Montreal benchmarks demonstrate that MobEvolve significantly outperforms state-of-the-art deep generative and LLM-based methods in individual trajectory fidelity, population-level distribution alignment, and behavioral plausibility, while preserving interpretability and high inference efficiency.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper introduces MobEvolve, the first agentic self-evolving heuristic framework for human mobility generation. It initializes a behavior-inspired heuristic system and employs an LLM agent to iteratively evolve its internal logic by diagnosing empirical misalignments and failure cases on a validation set, proposing targeted updates, and accumulating evolution memory for cumulative self-improvement. Extensive evaluations on the Singapore and Montreal benchmarks are claimed to demonstrate that MobEvolve significantly outperforms state-of-the-art deep generative and LLM-based methods in individual trajectory fidelity, population-level distribution alignment, and behavioral plausibility, while preserving interpretability and high inference efficiency.

Significance. If the claimed outperformance and robustness of the self-evolution process hold, the work could be significant by offering an interpretable, adaptive alternative to opaque deep generative models in human mobility synthesis. The integration of LLM-driven heuristic evolution with validation-set diagnostics represents a potentially useful paradigm for balancing performance, plausibility, and explainability in trajectory generation tasks.

major comments (2)
  1. [Abstract] Abstract: the claim that MobEvolve 'significantly outperforms' SOTA methods on the Singapore and Montreal benchmarks is asserted without any quantitative results, error bars, statistical tests, description of the initial heuristics, or the precise update mechanism; this makes the central empirical claim impossible to evaluate.
  2. [Abstract] Abstract: the load-bearing assumption that the LLM agent reliably diagnoses misalignments and proposes non-regressive updates without hallucinations, bias introduction, or validation-set overfitting is unsupported by any mentioned safeguards such as formal verification of updates, multi-agent consensus, or post-update ablation on held-out data.
minor comments (1)
  1. [Abstract] The abstract would benefit from a concise statement of the key metrics used for fidelity, alignment, and plausibility to aid immediate assessment of the claimed improvements.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thoughtful and constructive comments. We agree that the abstract would benefit from greater specificity to make the empirical claims more evaluable and will revise it in the next version. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that MobEvolve 'significantly outperforms' SOTA methods on the Singapore and Montreal benchmarks is asserted without any quantitative results, error bars, statistical tests, description of the initial heuristics, or the precise update mechanism; this makes the central empirical claim impossible to evaluate.

    Authors: We acknowledge this limitation in the current abstract. The full manuscript reports quantitative results with means and standard deviations (error bars) across 5 runs in Tables 2 and 3, along with paired t-tests for statistical significance (p<0.01) against baselines. Initial heuristics are detailed in Section 3.1 (behavior-inspired rules for trip chaining) and the update mechanism in Section 3.2 (LLM-proposed edits to heuristic logic with evolution memory). To address the concern directly, we will revise the abstract to incorporate 1-2 representative quantitative improvements (e.g., 'achieves 18% higher individual fidelity and 12% better population alignment') and a concise description of heuristic initialization and the targeted update process, subject to length constraints. revision: yes

  2. Referee: [Abstract] Abstract: the load-bearing assumption that the LLM agent reliably diagnoses misalignments and proposes non-regressive updates without hallucinations, bias introduction, or validation-set overfitting is unsupported by any mentioned safeguards such as formal verification of updates, multi-agent consensus, or post-update ablation on held-out data.

    Authors: This is a valid critique of the current presentation. The manuscript relies on empirical checks via repeated validation-set evaluation and evolution memory to track cumulative improvements, but does not explicitly describe additional safeguards in the abstract or main text. In revision we will add a dedicated subsection (likely in Section 3 or 5) that (a) reports post-update ablation results on a held-out test set to verify non-regression, (b) notes the use of prompt engineering and manual inspection of proposed updates to reduce hallucination risk, and (c) discusses the absence of multi-agent consensus as a current design choice while showing that single-agent updates have not introduced measurable bias in our experiments. These additions will be supported by new ablation figures without changing the core method. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical system evaluated on external benchmarks with no fitted predictions or self-referential derivations

full rationale

The paper describes an agentic framework that initializes behavior-inspired heuristics and uses an LLM agent to propose iterative updates based on validation-set diagnostics, with final performance measured via direct comparison to SOTA methods on the Singapore and Montreal benchmarks. No equations, parameter fits, or first-principles derivations are presented; the claimed improvements are reported as external empirical outcomes rather than quantities forced by construction from the inputs. The central assumption about LLM reliability is an empirical risk, not a circular reduction. This matches the default expectation of a non-circular system paper.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no internal equations, parameters, or modeling assumptions are described, so the ledger cannot be populated with concrete entries.

pith-pipeline@v0.9.1-grok · 5717 in / 1033 out tokens · 20314 ms · 2026-06-28T14:45:00.795174+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

96 extracted references · 1 canonical work pages

  1. [1]

    Lennart Adenaw and Quirin Bachmeier. 2022. Generating activity-based mobility plans from trip-based models and mobility surveys. Applied Sciences, 12(17):8456

    2022

  2. [2]

    Licia Amichi, Joon-Seok Kim, Gautam Malviya Thakur, and Carter Christopher. 2025. Exploring the utility-privacy trade-off: Impacts of semantic and visit types ambiguities on human mobility simulation. In 2025 26th IEEE International Conference on Mobile Data Management (MDM), pages 90--95. IEEE

    2025

  3. [3]

    Anthropic . 2026. Claude Code . https://docs.anthropic.com/en/docs/agents-and-tools/claude-code/overview. AI-powered coding assistant. Accessed: 2026-05-25

    2026

  4. [4]

    Theo Arentze, Frank Hofman, Henk Van Mourik, and Harry Timmermans. 2000. Albatross: multiagent, rule-based model of activity pattern decisions. Transportation Research Record, 1706(1):136--144

    2000

  5. [5]

    Theo A Arentze and Harry JP Timmermans. 2004. A learning-based transportation oriented simulation system. Transportation Research Part B: Methodological, 38(7):613--633

    2004

  6. [6]

    Hugo Barbosa, Marc Barthelemy, Gourab Ghoshal, Charlotte R James, Maxime Lenormand, Thomas Louail, Ronaldo Menezes, Jos \'e J Ramasco, Filippo Simini, and Marcello Tomasini. 2018. Human mobility: Models and applications. Physics Reports, 734:1--74

    2018

  7. [7]

    Richard J Beckman, Keith A Baggerly, and Michael D McKay. 1996. Creating synthetic baseline populations. Transportation Research Part A: Policy and Practice, 30(6):415--429

    1996

  8. [8]

    Jo \ a o Carlos N Bittencourt, Thommas KS Flores, Thiago C Jesus, and Daniel G Costa. 2026. On the role of ai in building generative urban intelligence. Artificial Intelligence Review, 59(2):78

    2026

  9. [9]

    John L Bowman and Moshe E Ben-Akiva. 2001. Activity-based disaggregate travel demand model system with activity schedules. Transportation research part a: policy and practice, 35(1):1--28

    2001

  10. [10]

    Dirk Brockmann, Lars Hufnagel, and Theo Geisel. 2006. The scaling laws of human travel. Nature, 439(7075):462--465

    2006

  11. [12]

    Brian d'Alessandro, Cathy O'Neil, and Tom LaGatta. 2017. Conscientious classification: A data scientist's guide to discrimination-aware classification. Big data, 5(2):120--134

    2017

  12. [13]

    Yves-Alexandre De Montjoye, C \'e sar A Hidalgo, Michel Verleysen, and Vincent D Blondel. 2013. Unique in the crowd: The privacy bounds of human mobility. Scientific reports, 3(1):1376

    2013

  13. [14]

    Nathan Eagle and Alex Sandy Pentland. 2009. Eigenbehaviors: Identifying structure in routine. Behavioral ecology and sociobiology, 63(7):1057--1066

    2009

  14. [15]

    Marta C Gonzalez, Cesar A Hidalgo, and Albert-Laszlo Barabasi. 2008. Understanding individual human mobility patterns. nature, 453(7196):779--782

    2008

  15. [16]

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2020. Generative adversarial networks. Communications of the ACM, 63(11):139--144

    2020

  16. [17]

    Baoshen Guo, Zhiqing Hong, Lidan Cao, Donghang Li, Junyi Li, Can Rong, Alok Prakash, Shenhao Wang, and Jinhua Zhao. 2025. Language models meet urban mobility: A data-centric review. Authorea Preprints

    2025

  17. [18]

    Alexandra Kapp, Julia Hansmeyer, and Helena Mihaljevi \'c . 2023. Generative models for synthetic urban mobility data: A systematic literature review. ACM Computing Surveys, 56(4):1--37

    2023

  18. [20]

    Ryuichi Kitamura, Cynthia Chen, and Ram M Pendyala. 1997. Generation of synthetic daily activity-travel patterns. Transportation research record, 1607(1):154--162

    1997

  19. [21]

    Chanyeong Kwak and Alan Clayton-Matthews. 2002. Multinomial logistic regression. Nursing research, 51(6):404--410

    2002

  20. [23]

    Yuebing Liang, Shenhao Wang, Jiangbo Yu, Zhan Zhao, Jinhua Zhao, and Sandy Pentland. 2026. Analyzing sequential activity and travel decisions with interpretable deep inverse reinforcement learning. Travel Behaviour and Society, 43:101171

    2026

  21. [24]

    Qingyue Long, Yuan Yuan, and Yong Li. 2025. A universal model for human mobility prediction. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 1, pages 894--905

    2025

  22. [25]

    Massimiliano Luca, Gianni Barlacchi, Bruno Lepri, and Luca Pappalardo. 2021. A survey on deep learning for human mobility. ACM Computing Surveys (CSUR), 55(1):1--44

    2021

  23. [26]

    Yecheng Jason Ma, William Liang, Guanzhi Wang, De-An Huang, Osbert Bastani, Dinesh Jayaraman, Yuke Zhu, Jim Fan, and 1 others. 2024. Eureka: Human-level reward design via coding large language models. In International conference on learning Representations, volume 2024, pages 26516--26560

    2024

  24. [27]

    Ngoc An Nguyen, Joerg Schweizer, and Federico Rupi. 2025. Large-scale activity-based demand generation modeling: A literature review and exploration of potential approaches. Transportation Engineering, 20:100329

    2025

  25. [28]

    James R Norris. 1998. Markov chains. 2. Cambridge university press

    1998

  26. [30]

    OpenAI . 2026. Codex . https://developers.openai.com/codex. OpenAI coding agent for software development. Accessed: 2026-05-25

    2026

  27. [31]

    Luca Pappalardo and Filippo Simini. 2018. Data-driven generation of spatio-temporal routines in human mobility. Data Mining and Knowledge Discovery, 32(3):787--829

    2018

  28. [33]

    Neha Patki, Roy Wedge, and Kalyan Veeramachaneni. 2016 b . The synthetic data vault. In 2016 IEEE international conference on data science and advanced analytics (DSAA), pages 399--410. IEEE

    2016

  29. [35]

    Negar Rezvany, Marija Kukic, and Michel Bierlaire. 2024. A review of activity-based disaggregate travel demand models. Findings

    2024

  30. [36]

    Bernardino Romera-Paredes, Mohammadamin Barekatain, Alexander Novikov, Matej Balog, M Pawan Kumar, Emilien Dupont, Francisco JR Ruiz, Jordan S Ellenberg, Pengming Wang, Omar Fawzi, and 1 others. 2024. Mathematical discoveries from program search with large language models. Nature, 625(7995):468--475

    2024

  31. [37]

    Debora Russo, Franca Rocco Di Torrepadula, Luigi Libero Lucio Starace, Sergio Di Martino, and Nicola Mazzocca. 2025. A framework for generating synthetic urban mobility datasets with customizable anomalous scenarios. IEEE Open Journal of Intelligent Transportation Systems, 6:1439--1458

    2025

  32. [38]

    Filippo Simini, Marta C Gonz \'a lez, Amos Maritan, and Albert-L \'a szl \'o Barab \'a si. 2012. A universal model for mobility and migration patterns. Nature, 484(7392):96--100

    2012

  33. [39]

    M Sklar. 1959. Fonctions de r \'e partition \`a n dimensions et leurs marges. In Annales de l'ISUP, volume 8, pages 229--231

    1959

  34. [40]

    Chaoming Song, Zehui Qu, Nicholas Blumm, and Albert-L \'a szl \'o Barab \'a si. 2010. Limits of predictability in human mobility. Science, 327(5968):1018--1021

    2010

  35. [42]

    Yihong Tang, Zhaokai Wang, Ao Qu, Yihao Yan, Zhaofeng Wu, Dingyi Zhuang, Jushi Kai, Kebing Hou, Xiaotong Guo, Jinhua Zhao, and 1 others. 2024. Itinera: Integrating spatial optimization with large language models for open-domain urban itinerary planning. In Proceedings of the 2024 conference on empirical methods in natural language processing: Industry tra...

    2024

  36. [43]

    Syed Mahir Tazwar, Max Knobbout, Enrique Hortal Quesada, and Mirela Popa. 2024. Tab-vae: A novel vae for generating synthetic tabular data. In ICPRAM, pages 17--26

    2024

  37. [44]

    Kay W Axhausen, Andreas Horni, and Kai Nagel. 2016. The multi-agent transport simulation MATSim. Ubiquity Press

    2016

  38. [45]

    Jiawei Wang, Renhe Jiang, Chuang Yang, Zengqing Wu, Makoto Onizuka, Ryosuke Shibasaki, Noboru Koshizuka, and Chuan Xiao. 2024. Large language models as urban residents: An llm agent framework for personal mobility generation. Advances in Neural Information Processing Systems, 37:124547--124574

    2024

  39. [46]

    Zi Wang and Fang Ren. 2025. Developing a decision support system for sustainable urban planning using machine learning-based scenario modeling. Scientific Reports, 15(1):13210

    2025

  40. [47]

    Jiayi Weng. 2026. Learning beyond gradients. https://trinkle23897.github.io/learning-beyond-gradients/. Blog post

    2026

  41. [48]

    Xiangping Wu, Zheng Zhang, Wangjun Wan, and Shuaiwei Yao. 2024. Personalized behavior modeling network for human mobility prediction. Journal of Ambient Intelligence and Humanized Computing, 15(9):3289--3301

    2024

  42. [50]

    Lei Xu, Maria Skoularidou, Alfredo Cuesta-Infante, and Kalyan Veeramachaneni. 2019. Modeling tabular data using conditional gan. Advances in neural information processing systems, 32

    2019

  43. [51]

    Hongtao Zhang and Lingcheng Dai. 2018. Mobility prediction: A survey on state-of-the-art schemes and future applications. IEEE access, 7:802--822

    2018

  44. [53]

    arXiv preprint arXiv:2505.14752 , year=

    LLMSynthor: Macro-Aligned Micro-Records Synthesis with Large Language Models , author=. arXiv preprint arXiv:2505.14752 , year=

    Pith/arXiv arXiv

  45. [54]

    Nature , volume=

    The scaling laws of human travel , author=. Nature , volume=. 2006 , publisher=

    2006

  46. [55]

    2016 IEEE international conference on data science and advanced analytics (DSAA) , pages=

    The synthetic data vault , author=. 2016 IEEE international conference on data science and advanced analytics (DSAA) , pages=. 2016 , organization=

    2016

  47. [56]

    Fonctions de r

    Sklar, M , booktitle=. Fonctions de r

  48. [57]

    Applied Sciences , volume=

    Generating activity-based mobility plans from trip-based models and mobility surveys , author=. Applied Sciences , volume=. 2022 , publisher=

    2022

  49. [58]

    Transportation Engineering , volume=

    Large-scale activity-based demand generation modeling: A literature review and exploration of potential approaches , author=. Transportation Engineering , volume=. 2025 , publisher=

    2025

  50. [59]

    Findings , year=

    A review of activity-based disaggregate travel demand models , author=. Findings , year=

  51. [60]

    Transportation research record , volume=

    Generation of synthetic daily activity-travel patterns , author=. Transportation research record , volume=. 1997 , publisher=

    1997

  52. [61]

    arXiv preprint arXiv:2109.02715 , year=

    Individual mobility prediction via attentive marked temporal point processes , author=. arXiv preprint arXiv:2109.02715 , year=

    arXiv

  53. [62]

    IEEE Transactions on Intelligent Transportation Systems , volume=

    Activity-aware human mobility prediction with hierarchical graph attention recurrent network , author=. IEEE Transactions on Intelligent Transportation Systems , volume=. 2024 , publisher=

    2024

  54. [63]

    arXiv preprint arXiv:1312.6114 , year=

    Auto-encoding variational bayes , author=. arXiv preprint arXiv:1312.6114 , year=

    Pith/arXiv arXiv

  55. [64]

    Communications of the ACM , volume=

    Generative adversarial networks , author=. Communications of the ACM , volume=. 2020 , publisher=

    2020

  56. [65]

    Nature , volume=

    A universal model for mobility and migration patterns , author=. Nature , volume=. 2012 , publisher=

    2012

  57. [66]

    Physics Reports , volume=

    Human mobility: Models and applications , author=. Physics Reports , volume=. 2018 , publisher=

    2018

  58. [67]

    IEEE access , volume=

    Mobility prediction: A survey on state-of-the-art schemes and future applications , author=. IEEE access , volume=. 2018 , publisher=

    2018

  59. [68]

    Scientific Reports , volume=

    Developing a decision support system for sustainable urban planning using machine learning-based scenario modeling , author=. Scientific Reports , volume=. 2025 , publisher=

    2025

  60. [69]

    Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V

    A universal model for human mobility prediction , author=. Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 1 , pages=

  61. [70]

    IEEE Open Journal of Intelligent Transportation Systems , volume=

    A Framework for Generating Synthetic Urban Mobility Datasets With Customizable Anomalous Scenarios , author=. IEEE Open Journal of Intelligent Transportation Systems , volume=. 2025 , publisher=

    2025

  62. [71]

    Scientific reports , volume=

    Unique in the crowd: The privacy bounds of human mobility , author=. Scientific reports , volume=. 2013 , publisher=

    2013

  63. [72]

    Transportation Research Record , volume=

    ALBATROSS: multiagent, rule-based model of activity pattern decisions , author=. Transportation Research Record , volume=. 2000 , publisher=

    2000

  64. [73]

    nature , volume=

    Understanding individual human mobility patterns , author=. nature , volume=. 2008 , publisher=

    2008

  65. [74]

    2025 26th IEEE International Conference on Mobile Data Management (MDM) , pages=

    Exploring the Utility-Privacy Trade-Off: Impacts of Semantic and Visit Types Ambiguities on Human Mobility Simulation , author=. 2025 26th IEEE International Conference on Mobile Data Management (MDM) , pages=. 2025 , organization=

    2025

  66. [75]

    Science , volume=

    Limits of predictability in human mobility , author=. Science , volume=. 2010 , publisher=

    2010

  67. [76]

    Behavioral ecology and sociobiology , volume=

    Eigenbehaviors: Identifying structure in routine , author=. Behavioral ecology and sociobiology , volume=. 2009 , publisher=

    2009

  68. [77]

    Data Mining and Knowledge Discovery , volume=

    Data-driven generation of spatio-temporal routines in human mobility , author=. Data Mining and Knowledge Discovery , volume=. 2018 , publisher=

    2018

  69. [78]

    Patki, R

    The Synthetic data vault , author=. IEEE International Conference on Data Science and Advanced Analytics (DSAA) , year=. doi:10.1109/DSAA.2016.49 , month=

    work page doi:10.1109/dsaa.2016.49 2016

  70. [79]

    Transportation research part a: policy and practice , volume=

    Activity-based disaggregate travel demand model system with activity schedules , author=. Transportation research part a: policy and practice , volume=. 2001 , publisher=

    2001

  71. [80]

    Advances in neural information processing systems , volume=

    Modeling tabular data using conditional gan , author=. Advances in neural information processing systems , volume=

  72. [81]

    2016 , publisher=

    The multi-agent transport simulation MATSim , author=. 2016 , publisher=

    2016

  73. [82]

    Transportation Research Part A: Policy and Practice , volume=

    Creating synthetic baseline populations , author=. Transportation Research Part A: Policy and Practice , volume=. 1996 , publisher=

    1996

  74. [83]

    , author=

    Tab-VAE: A Novel VAE for Generating Synthetic Tabular Data. , author=. ICPRAM , pages=

  75. [84]

    Big data , volume=

    Conscientious classification: A data scientist's guide to discrimination-aware classification , author=. Big data , volume=. 2017 , publisher=

    2017

  76. [85]

    Transportation Research Part B: Methodological , volume=

    A learning-based transportation oriented simulation system , author=. Transportation Research Part B: Methodological , volume=. 2004 , publisher=

    2004

  77. [86]

    ACM Computing Surveys , volume=

    Generative models for synthetic urban mobility data: A systematic literature review , author=. ACM Computing Surveys , volume=. 2023 , publisher=

    2023

  78. [87]

    Artificial Intelligence Review , volume=

    On the role of AI in building generative urban intelligence , author=. Artificial Intelligence Review , volume=. 2026 , publisher=

    2026

  79. [88]

    Authorea Preprints , year=

    Language Models Meet Urban Mobility: A Data-Centric Review , author=. Authorea Preprints , year=

  80. [89]

    arXiv preprint arXiv:2503.07158 , year=

    Generative ai in transportation planning: A survey , author=. arXiv preprint arXiv:2503.07158 , year=

    arXiv

Showing first 80 references.