Who you are determines how you travel: Clustering human activity patterns with a Markov-chain-based mixture model

Zhou, Yang; Yuan, Quan; Yang, Chao; Wang, Yinhai

doi:10.1016/j.tbs.2021.03.005

Cited by 25 publications

(8 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Learned representation based methods: The trajectory is first represented by a sequence and then encoded into a vector representation. The similarity between trajectory is calculated by vector similarity [5,10,12,20]- [22,24]. Markov chain based method [12,24] and recurrent neural networks (RNNs) [5,20]- [22] are the commonly used components.…”

Section: Revisit Existing Trajectory Representationmentioning

confidence: 99%

“…The similarity between trajectory is calculated by vector similarity [5,10,12,20]- [22,24]. Markov chain based method [12,24] and recurrent neural networks (RNNs) [5,20]- [22] are the commonly used components.…”

Section: Revisit Existing Trajectory Representationmentioning

confidence: 99%

“…Understanding individual human mobility patterns has been an actively-studied topic in the past decade [7,8,14,24]. Clustering human mobility is an important task to group "similar" mobility behaviors and provides insight towards detected mobility behaviors patterns.…”

Section: Introductionmentioning

confidence: 99%

“…Existing trajectory clustering methods follow a paradigm that first chooses a way to represent the trajectory and then applies a clustering algorithm on the trajectory representation to group similar trajectories [8,14,19]- [22,24]. Although many advances have been achieved, there are still major challenges.…”

Section: Introductionmentioning

confidence: 99%

“…First, even though there are different ways to represent a trajectory, they fail to properly handle the diversity of multi-scale trajectories within the same mobility type. The core idea of the existing methods is either directly representing a trajectory as a temporal sequence of which the order between consecutive elements (e.g., recorded geocoordinates or stay points) represents the transition between corresponding consecutive time steps in the trajectory [2,9] or learning a vector representation encoding all the elements and transitions from the temporal sequence [20]- [22,24]. The definition of an element varies across methods but the transition direction always represents the temporal order.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Clustering Human Mobility with Multiple Spaces

Lin

Chiang

2022

2022 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

Human mobility clustering is an important problem for understanding human mobility behaviors (e.g., work and school commutes). Existing methods typically contain two steps: choosing/learning a mobility representation and applying a clustering algorithm to the representation. However, these methods rely on strict visiting orders in trajectories and cannot take advantage of multiple types of mobility representations. This paper proposes a novel mobility clustering method for mobility behavior detection. First, the proposed method contains a permutation-equivalent operation to handle sub-trajectories that might have different visiting orders but similar impacts on mobility behaviors. Second, the proposed method utilizes a variational autoencoder architecture to simultaneously perform clustering in both latent and original spaces. Also, in order to handle the bias of a single latent space, our clustering assignment prediction considers multiple learned latent spaces at different epochs. This way, the proposed method produces accurate results and can provide reliability estimates of each trajectory's cluster assignment. The experiment shows that the proposed method outperformed state-of-the-art methods in mobility behavior detection from trajectories with better accuracy and more interpretability.

show abstract