Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

PateGail: A Privacy-Preserving Mobility Trajectory Generator with Imitation Learning (2407.16729v1)

Published 23 Jul 2024 in cs.LG and cs.AI

Abstract: Generating human mobility trajectories is of great importance to solve the lack of large-scale trajectory data in numerous applications, which is caused by privacy concerns. However, existing mobility trajectory generation methods still require real-world human trajectories centrally collected as the training data, where there exists an inescapable risk of privacy leakage. To overcome this limitation, in this paper, we propose PateGail, a privacy-preserving imitation learning model to generate mobility trajectories, which utilizes the powerful generative adversary imitation learning model to simulate the decision-making process of humans. Further, in order to protect user privacy, we train this model collectively based on decentralized mobility data stored in user devices, where personal discriminators are trained locally to distinguish and reward the real and generated human trajectories. In the training process, only the generated trajectories and their rewards obtained based on personal discriminators are shared between the server and devices, whose privacy is further preserved by our proposed perturbation mechanisms with theoretical proof to satisfy differential privacy. Further, to better model the human decision-making process, we propose a novel aggregation mechanism of the rewards obtained from personal discriminators. We theoretically prove that under the reward obtained based on the aggregation mechanism, our proposed model maximizes the lower bound of the discounted total rewards of users. Extensive experiments show that the trajectories generated by our model are able to resemble real-world trajectories in terms of five key statistical metrics, outperforming state-of-the-art algorithms by over 48.03%. Furthermore, we demonstrate that the synthetic trajectories are able to efficiently support practical applications, including mobility prediction and location recommendation.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (47)
  1. Social GAN: Socially Acceptable Trajectories With Generative Adversarial Networks. In Proc. CVPR.
  2. A Framework for Behavioural Cloning. In Machine Intelligence 15, 103–129.
  3. Synthesizing plausible privacy-preserving location traces. In Proc. IEEE SP.
  4. Relative entropy inverse reinforcement learning. In Proc. AISTATS, 182–189.
  5. Federated meta-learning with fast convergence and efficient communication. arXiv preprint arXiv:1802.07876.
  6. The algorithmic foundations of differential privacy. Found. Trends Theor. Comput. Sci., 9(3-4): 211–407.
  7. AutoMEC: LSTM-based user mobility prediction for service management in distributed MEC resources. In Proc. MSWiM.
  8. Identification of critical roads in urban transportation network based on GPS trajectory data. Physica A: Statistical Mechanics and its Applications, 535: 122337.
  9. Deepmove: Predicting human mobility with attentional recurrent networks. In Proc. WWW.
  10. Learning to Simulate Human Mobility. Proc. KDD.
  11. Inverting gradients-how easy is it to break privacy in federated learning? Proc. NeurIPS.
  12. Understanding individual human mobility patterns. Nature, 453(7196): p.779–782.
  13. Generative adversarial nets. In Proc. NeurIPS.
  14. Federated learning for mobile keyboard prediction. arXiv preprint arXiv:1811.03604.
  15. Fast matrix factorization for online recommendation with implicit feedback. In Proc. SIGIR.
  16. Data-driven Human Mobility Modeling: A Survey and Engineering Guidance for Mobile Networking. ACM Computing Surveys (CSUR).
  17. Generative adversarial imitation learning. In Proc. NeurIPS.
  18. A variational autoencoder based generative model of urban human mobility. In Proc. MIPR.
  19. Human mobility modeling at metropolitan scales. 239–252.
  20. The TimeGeo modeling framework for urban mobility without travel surveys. PNAS, 113(37).
  21. PATE-GAN: Generating synthetic data with differential privacy guarantees. In Proc. ICLR.
  22. Generative models for simulating mobility trajectories. arXiv preprint arXiv:1811.12801.
  23. trajGANs : Using generative adversarial networks for geo-privacy protection of trajectory data ( Vision paper ). In Location Privacy and Security Workshop.
  24. “Is Reinforcement Learning the Choice of Human Learners?: A Case Study of Taxi Drivers”. In Proc. SIGSPATIAL.
  25. Asynchronous methods for deep reinforcement learning. In Proc. ICML.
  26. Comprehensive privacy analysis of deep learning: Passive and active white-box inference attacks against centralized and federated learning. In Proc. IEEE SP, 739–753.
  27. A Non-Parametric Generative Model for Human Trajectories. In IJCAI, 3812–3817.
  28. xGAIL: Explainable Generative Adversarial Imitation Learning for Explainable Human Decision Analysis. In Proc. KDD.
  29. Semi-supervised knowledge transfer for deep learning from private training data. arXiv preprint arXiv:1610.05755.
  30. Data-driven generation of spatio-temporal routines in human mobility. Data Mining and Knowledge Discovery, 32(1).
  31. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347.
  32. “TrajGAIL: Generating Urban Vehicle Trajectories using Generative Adversarial Imitation Learning”. In arXiv preprint.
  33. Membership Inference Attacks Against Machine Learning Models. 3–18.
  34. Limits of Predictability in Human Mobility. Science, 327(5968): 1018–1021.
  35. Elements of information theory. John Wiley & Sons.
  36. Attention is all you need. In Proc. NeurIPS.
  37. Adversarial Cooperative Imitation Learning for Dynamic Treatment Regimes. In Proc. WWW.
  38. A Joint Inverse Reinforcement Learning and Deep Learning Model for Drivers’ Behavioral Prediction. In Proc. CIKM.
  39. Learning Traffic Signal Control from Demonstrations. In Proc. CIKM.
  40. A generative model of urban activities from cellular data. IEEE Transactions on Intelligent Transportation Systems, 19(6).
  41. Seqgan: Sequence generative adversarial nets with policy gradient. In Proc. AAAI.
  42. TrajGAIL: Trajectory Generative Adversarial Imitation Learning for Long-term Decision Analysis. In Proc. ICDM.
  43. Unveiling taxi drivers’ strategies via cgail: Conditional generative adversarial imitation learning. In Proc. ICDM.
  44. Synthesizing Privacy Preserving Traces: Enhancing Plausibility With Social Networks. IEEE/ACM Transactions on Networking, PP(99): 1–14.
  45. Geolife: A collaborative social networking service among user, location and trajectory. IEEE Data Eng. Bull., 33(2): 32–39.
  46. Modeling interaction via the principle of maximum causal entropy. In Proc. ICML.
  47. Maximum entropy inverse reinforcement learning. In Proc. AAAI.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Huandong Wang (35 papers)
  2. Changzheng Gao (1 paper)
  3. Yuchen Wu (42 papers)
  4. Depeng Jin (72 papers)
  5. Lina Yao (194 papers)
  6. Yong Li (628 papers)
Citations (13)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets