论文标题
Totopo：将单变量和多变量时间序列分类与拓扑数据分析
TOTOPO: Classifying univariate and multivariate time series with Topological Data Analysis
论文作者
论文摘要
这项工作致力于对拓扑数据分析Fortime Series Clastification的全面分析。先前的工作有重大的缺点，例如大规模的基准测试或缺少最新方法。在这项工作中，我们提出了用于从不同类型的垂直图图中提取拓扑描述符的Totopo。结果表明，托托波（Totopo）在准确性方面显着超过了基线。 Totopo也与现有的Thestate具有竞争力，在20％的单变量和40％的多元时间表数据集中是最好的。这项工作验证了以下假设：基于TDA的方法对数据中的小扰动进行了敏捷，并且对于周期性和形象有助于区分班级的情况很有用。
This work is devoted to a comprehensive analysis of topological data analysis fortime series classification. Previous works have significant shortcomings, such aslack of large-scale benchmarking or missing state-of-the-art methods. In this work,we propose TOTOPO for extracting topological descriptors from different types ofpersistence diagrams. The results suggest that TOTOPO significantly outperformsexisting baselines in terms of accuracy. TOTOPO is also competitive with thestate-of-the-art, being the best on 20% of univariate and 40% of multivariate timeseries datasets. This work validates the hypothesis that TDA-based approaches arerobust to small perturbations in data and are useful for cases where periodicity andshape help discriminate between classes.
