Using Relative Distance and Hausdorff Distance to Mine Trajectory Clusters

Guan, Bo; Liu, Liangxu; Chen, Jinyang

doi:10.11591/telkomnika.v11i1.1877

Cited by 13 publications

(11 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Final detection of anomalous sub-trajectories was done using a hybrid of the distance-based and density-based approaches. Guan et al [20] defined the distance between sub-trajectories based on local and relative distance from the line Hausdorff distance. In addition, an R-tree was employed to improve the efficiency of the DBSCAN method.…”

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

“…There have been many methods used to realize a similarity measurement, including Euclidean distance [23], Hausdorff distance [20], LCSS (longest common subsequences) [24], DTW (dynamic time warping) [7], ERP (edit distance with real penalty) [25], and EDR (edit distance in real sequence) [26]. In practical application, the Euclidean distance requires two trajectories of the same length.…”

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

“…The selection of a suitable clustering algorithm includes a density-based method [20,23,29], spectral clustering [30], a model-based method [12], or hierarchical clustering [31]. Lee et al [23] chose a density-based line-segment clustering algorithm for grouping similar line segments together.…”

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

“…Bermingham et al [29] improved TRACLUS to create ND-TRACLUS, which introduced Retraspam, and split and merged the most representative line segments simply. Guan et al [20] proposed trajectory clustering based on an improved minimum Hausdorff distance, named TraClustMHD. Fu et al [30] first resampled trajectories by equal space intervals.…”

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

See 3 more Smart Citations

Detecting Anomalous Trajectories and Behavior Patterns Using Hierarchical Clustering from Taxi GPS Data

Wang

Qin

Chen

et al. 2018

IJGI

View full text Add to dashboard Cite

Anomalous taxi trajectories are those chosen by a small number of drivers that are different from the regular choices of other drivers. These anomalous driving trajectories provide us an opportunity to extract driver or passenger behaviors and monitor adverse urban traffic events. Because various trajectory clustering methods have previously proven to be an effective means to analyze similarities and anomalies within taxi GPS trajectory data, we focus on the problem of detecting anomalous taxi trajectories, and we develop our trajectory clustering method based on the edit distance and hierarchical clustering. To achieve this objective, first, we obtain all the taxi trajectories crossing the same source-destination pairs from taxi trajectories and take these trajectories as clustering objects. Second, an edit distance algorithm is modified to measure the similarity of the trajectories. Then, we distinguish regular trajectories and anomalous trajectories by applying adaptive hierarchical clustering based on an optimal number of clusters. Moreover, we further analyze these anomalous trajectories and discover four anomalous behavior patterns to speculate on the cause of an anomaly based on statistical indicators of time and length. The experimental results show that the proposed method can effectively detect anomalous trajectories and can be used to infer clearly fraudulent driving routes and the occurrence of adverse traffic events.

show abstract

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

Section: Related Work About Trajectory Clustering For Anomalous Trajmentioning

confidence: 99%

See 2 more Smart Citations

Detecting Anomalous Trajectories and Behavior Patterns Using Hierarchical Clustering from Taxi GPS Data

Wang

Qin

Chen

et al. 2018

IJGI

View full text Add to dashboard Cite

show abstract

“…Track Registration means to restore or repair the loss shape of track caused by collection frequency and vehicle speed. There are two kinds of track registration method, one is Map Matching, and another is Multi-track Fusion [5], [6].…”

Section: Introductionmentioning

confidence: 99%

A method of Trajectory Restoration at Intersection

Zheng

Song²,

Zhu³

2015

TELKOMNIKA

View full text Add to dashboard Cite

show abstract

Ecological dynamic regimes: Identification, characterization, and comparison

Sánchez‐Pinillos

Kéfi

Cáceres

et al. 2023

Ecological Monographs

View full text Add to dashboard Cite

Understanding ecological dynamics has been a central topic in ecology since its origins. Yet, identifying dynamic regimes remains a research frontier for modern ecology. The concept of ecological dynamic regime emerged to emphasize the dynamic property of steady states in nature and refers to the fluctuations of ecosystems around some trend or average. Identifying and characterizing ecological dynamic regimes is of utmost importance in the current context of global change since they form the reference against which post‐disturbance dynamics must be compared to assess ecological resilience. However, the implementation of ecological dynamic regimes in empirical science is still challenging given the high dimensionality and stochasticity of ecological data and the large volume of data required to distinguish stochastic dynamics from general and predictable dynamics. The era of big data and the recent advances in quantitative ecology and data science offer an opportunity to study dynamic regimes using empirical approaches from a new perspective. This paper presents a novel methodological framework to describe ecological dynamic regimes from a set of ecological trajectories defined by the temporal changes of state variables in a multidimensional state space. In our framework, we formally define ecological dynamic regimes and include analytical tools to identify, characterize, and compare ecological dynamic regimes based on their geometric characteristics. More specifically, we propose different ways to identify ecological dynamic regimes from empirical data, develop a new algorithm to identify representative trajectories summarizing the main dynamic patterns, propose a set of metrics to describe the internal distribution of ecological trajectories, and define a dissimilarity index to compare two or more dynamic regimes based on their shape and position in the state space. We used artificial data to illustrate the different elements of our framework, and applied our analyses to real data, using permanent sampling plots of Canadian boreal forests as an example. Overall, our framework contributes to filling the gap between theoretical and empirical ecology by providing robust analytical tools to assess ecological resilience and study ecosystem dynamics from a multidimensional perspective and considering the variability of natural systems.

show abstract

Using Relative Distance and Hausdorff Distance to Mine Trajectory Clusters

Cited by 13 publications

References 12 publications

Detecting Anomalous Trajectories and Behavior Patterns Using Hierarchical Clustering from Taxi GPS Data

Detecting Anomalous Trajectories and Behavior Patterns Using Hierarchical Clustering from Taxi GPS Data

A method of Trajectory Restoration at Intersection

Ecological dynamic regimes: Identification, characterization, and comparison

Contact Info

Product

Resources

About