Station-to-User Transfer Learning: Towards Explainable User Clustering Through Latent Trip Signatures Using Tidal-Regularized Non-Negative Matrix Factorization

• URL: http://arxiv.org/abs/2004.12828v1
• Date: Mon, 27 Apr 2020 14:13:56 GMT
• Title: Station-to-User Transfer Learning: Towards Explainable User Clustering Through Latent Trip Signatures Using Tidal-Regularized Non-Negative Matrix Factorization
• Authors: Liming Zhang, Andreas Z\"ufle, Dieter Pfoser
• Abstract summary: This work focuses on mobility data and how it will help improve our understanding of urban mobility patterns. We propose a Collective Learning Framework through Latent Representation, which augments user-level learning with collective patterns learned from station-level signals. We provide a qualitative analysis of the station functions and user profiles for the Washington D.C. metro and show how our method supports intra-city mobility exploration.
• Score: 4.713006935605146
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Urban areas provide us with a treasure trove of available data capturing almost every aspect of a population's life. This work focuses on mobility data and how it will help improve our understanding of urban mobility patterns. Readily available and sizable farecard data captures trips in a public transportation network. However, such data typically lacks temporal modalities and as such the task of inferring trip semantic, station function, and user profile is quite challenging. As existing approaches either focus on station-level or user-level signals, they are prone to overfitting and generate less credible and insightful results. To properly learn such characteristics from trip data, we propose a Collective Learning Framework through Latent Representation, which augments user-level learning with collective patterns learned from station-level signals. This framework uses a novel, so-called Tidal-Regularized Non-negative Matrix Factorization method, which incorporates domain knowledge in the form of temporal passenger flow patterns in generic Non-negative Matrix Factorization. To evaluate our model performance, a user stability test based on the classical Rand Index is introduced as a metric to benchmark different unsupervised learning models. We provide a qualitative analysis of the station functions and user profiles for the Washington D.C. metro and show how our method supports spatiotemporal intra-city mobility exploration.

Related papers

• Unsupervised Domain Adaptation for Self-Driving from Past Traversal Features [69.47588461101925]
  We propose a method to adapt 3D object detectors to new driving environments. Our approach enhances LiDAR-based detection models using spatial quantized historical features. Experiments on real-world datasets demonstrate significant improvements.
  arXiv  Detail & Related papers  (2023-09-21T15:00:31Z)
• Clustering and Analysis of GPS Trajectory Data using Distance-based Features [20.91019606657394]
  We propose a new mobility metric, Daily Characteristic Distance, and use it to generate features for each user. We then use those features with an unsupervised machine learning method, $k$-means clustering, and obtain three clusters of users for each type of day (Workday and Offday) We propose two new metrics for the analysis of the clustering results, namely User Commonality and Average Frequency.
  arXiv  Detail & Related papers  (2022-12-01T01:25:49Z)
• Beyond Transfer Learning: Co-finetuning for Action Localisation [64.07196901012153]
  We propose co-finetuning -- simultaneously training a single model on multiple upstream'' and downstream'' tasks. We demonstrate that co-finetuning outperforms traditional transfer learning when using the same total amount of data. We also show how we can easily extend our approach to multiple upstream'' datasets to further improve performance.
  arXiv  Detail & Related papers  (2022-07-08T10:25:47Z)
• Spatio-Temporal Graph Few-Shot Learning with Cross-City Knowledge Transfer [58.6106391721944]
  Cross-city knowledge has shown its promise, where the model learned from data-sufficient cities is leveraged to benefit the learning process of data-scarce cities. We propose a model-agnostic few-shot learning framework for S-temporal graph called ST-GFSL. We conduct comprehensive experiments on four traffic speed prediction benchmarks and the results demonstrate the effectiveness of ST-GFSL compared with state-of-the-art methods.
  arXiv  Detail & Related papers  (2022-05-27T12:46:52Z)
• Statistical inference of travelers' route choice preferences with system-level data [4.120057972557892]
  We develop a methodology to estimate travelers' utility functions with multiple attributes using system-level data. Experiments on synthetic data show that the coefficients are consistently recovered and that hypothesis tests are a reliable statistic to identify which attributes are determinants of travelers' route choices. The methodology is also deployed at a large scale using real Fresnoworld multisource data collected during the COVID outbreak.
  arXiv  Detail & Related papers  (2022-04-23T00:38:32Z)
• Location Leakage in Federated Signal Maps [7.093808731951124]
  We consider the problem of predicting cellular network performance (signal maps) from measurements collected by several mobile devices. We formulate the problem within the online federated learning framework: (i) federated learning enables users to collaboratively train a model, while keeping their training data on their devices. We consider an honest-but-curious server, who observes the updates from target users participating in FL and infers their location using a deep leakage from gradients (DLG) type of attack. We build on this observation to protect location privacy, in our setting, by revisiting and designing mechanisms within the federated learning framework including: tuning the FL
  arXiv  Detail & Related papers  (2021-12-07T02:28:12Z)
• TraND: Transferable Neighborhood Discovery for Unsupervised Cross-domain Gait Recognition [77.77786072373942]
  This paper proposes a Transferable Neighborhood Discovery (TraND) framework to bridge the domain gap for unsupervised cross-domain gait recognition. We design an end-to-end trainable approach to automatically discover the confident neighborhoods of unlabeled samples in the latent space. Our method achieves state-of-the-art results on two public datasets, i.e., CASIA-B and OU-LP.
  arXiv  Detail & Related papers  (2021-02-09T03:07:07Z)
• AttnMove: History Enhanced Trajectory Recovery via Attentional Network [15.685998183691655]
  We propose a novel attentional neural network-based model, named AttnMove, to densify individual trajectories by recovering unobserved locations. We evaluate our model on two real-world datasets, and extensive results demonstrate the performance gain compared with the state-of-the-art methods.
  arXiv  Detail & Related papers  (2021-01-03T15:45:35Z)
• BREEDS: Benchmarks for Subpopulation Shift [98.90314444545204]
  We develop a methodology for assessing the robustness of models to subpopulation shift. We leverage the class structure underlying existing datasets to control the data subpopulations that comprise the training and test distributions. Applying this methodology to the ImageNet dataset, we create a suite of subpopulation shift benchmarks of varying granularity.
  arXiv  Detail & Related papers  (2020-08-11T17:04:47Z)
• Leveraging the Self-Transition Probability of Ordinal Pattern Transition Graph for Transportation Mode Classification [0.0]
  We propose the use of a feature retained from the Ordinal Pattern Transition Graph, called the probability of self-transition for transportation mode classification. The proposed feature presents better accuracy results than Permutation Entropy and Statistical Complexity, even when these two are combined.
  arXiv  Detail & Related papers  (2020-07-16T23:25:09Z)
• Learning Invariant Representations for Reinforcement Learning without Reconstruction [98.33235415273562]
  We study how representation learning can accelerate reinforcement learning from rich observations, such as images, without relying either on domain knowledge or pixel-reconstruction. Bisimulation metrics quantify behavioral similarity between states in continuous MDPs. We demonstrate the effectiveness of our method at disregarding task-irrelevant information using modified visual MuJoCo tasks.
  arXiv  Detail & Related papers  (2020-06-18T17:59:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.