Authors: Yaolin Liu*, Wuhan University, Yanfang Liu, Wuhan University
Topics: Geographic Information Science and Systems, Spatial Analysis & Modeling, Temporal GIS
Keywords: Clustering, time series, adaptive, data Mining.
Session Type: Paper
Start / End Time: 2:00 PM / 3:40 PM
Room: Grand Chenier, Sheraton, 5th Floor
Presentation File: No File Uploaded
Time series clustering models have been widely used to mine the clustering distribution characteristics of real phenomena. In the past few decades, many time series clustering models have been developed. These models can be roughly grouped into five classes, as follows: partitioning-based time series clustering models, hierarchical time series clustering models, density-based time series clustering models, graph-based time series clustering models and time series co-clustering models. Although these models can handle certain applications, they still suffer from several deficiencies and require improvement. In view of these shortcomings, an adaptive density-based time series clustering (DTSC) model has been proposed in this paper. DTSC model simultaneously considers the spatial attributes, non-spatial time series attribute values, and non-spatial time series attribute trends. DTSC models proceeds with two major parts. In the first part, the objects with spatial proximity relationship are considered as similar in the spatial domain. In the second part, an improved density-based clustering strategy is then adopted to detect clusters with similar non-spatial time series attribute values and time series attribute trends. The effectiveness and efficiency of the DTSC algorithm are validated by experiments on both time series rainfall dataset and time series surface deformation dataset. In the real applications, several interesting patterns have been found, and the results indicate that the proposed DTSC model can effectively detect time series clusters with arbitrary shapes and similar attributes and densities while considering noises