Mining latent patterns in geoMobile data via EPIC
Mining latent patterns in geoMobile data via EPIC
复制标题
DOI：
10.1007/s11280-019-00702-z
复制
发表时间：
2019-07
期刊：
影响因子：
--
通讯作者：
Arvind Narayanan;Saurabh Verma;Zhi-Li Zhang
中科院分区：
文献类型：
--
作者：
Arvind Narayanan;Saurabh Verma;Zhi-Li Zhang
我们创造了术语geoMobiledata，以强调表现出反映人类行为的地理空间特征的数据集。我们提出并开发了一个EPIC框架来从geoMobile数据中挖掘潜在模式，并提供有意义的解释：我们首先通过Laplacian Eigenmaps从高维geoMobile数据集中提取潜在特征，并在此潜在特征空间中进行聚类;然后使用最先进的可视化技术将这些潜在特征“P "投影到2D空间中;最后通过C 'ulling聚类特异性显著特征集得到有意义的I' interpretation。结果表明，该方法的局部空间收缩特性比其他主要的降维方法有上级的优越性。使用不同的现实世界的geoMobile数据集，我们通过三个案例研究表明我们的框架的有效性。
We coin the termgeoMobiledata to emphasize datasets that exhibit geo-spatial features reflective of human behaviors. We propose and develop anEPICframework to mine latent patterns from geoMobile data and provide meaningful interpretations: we first‘E’xtractlatent features from high dimensional geoMobile datasets via Laplacian Eigenmaps and perform clustering in this latent feature space; we then use a state-of-the-art visualization technique to‘P’rojectthese latent features into 2D space; and finally we obtain meaningful‘I’nterpretationsby‘C’ullingcluster-specific significant feature-set. We illustrate that the local space contraction property of our approach is most superior than other major dimension reduction techniques. Using diverse real-world geoMobile datasets, we show the efficacy of our framework via three case studies.