Di Iorio Jacopo, Cremona Marzia A, Chiaromonte Francesca
Department of Statistics, Penn State University, Joab L. Thomas Building, University Park, 16802 PA USA.
Department of Operations and Decision System, Université Laval, 2325 Rue de la Terrasse, Québec, G1V0A6 Québec Canada.
Stat Comput. 2025;35(1):11. doi: 10.1007/s11222-024-10537-y. Epub 2024 Dec 10.
Motif discovery is gaining increasing attention in the domain of functional data analysis. Functional motifs are typical "shapes" or "patterns" that recur multiple times in different portions of a single curve and/or in misaligned portions of multiple curves. In this paper, we define functional motifs using an additive model and we propose for their discovery and evaluation. Inspired by clustering and biclustering techniques, is a multi-step procedure which uses agglomerative hierarchical clustering with complete linkage and a functional distance based on mean squared residue scores to discover functional motifs, both in a single curve (e.g., time series) and in a set of curves. We assess its performance and compare it to other recent methods through extensive simulations. Moreover, we use for discovering motifs in two real-data case studies; one on food price inflation and one on temperature changes.
The online version contains supplementary material available at 10.1007/s11222-024-10537-y.
基序发现（Motif discovery）在功能数据分析领域正受到越来越多的关注。功能基序是指在单条曲线的不同部分和/或多条曲线的未对齐部分多次出现的典型“形状”或“模式”。在本文中，我们使用加法模型定义功能基序，并提出用于其发现和评估的方法。受聚类和双聚类技术的启发，该方法是一个多步骤过程，它使用具有完全链接的凝聚层次聚类和基于均方残差分数的功能距离来发现单条曲线（例如时间序列）和一组曲线中的功能基序。我们通过广泛的模拟评估其性能，并将其与其他近期方法进行比较。此外，我们在两个实际数据案例研究中使用该方法来发现基序；一个关于食品价格通胀，另一个关于温度变化。
在线版本包含可在10.1007/s11222-024-10537-y获取的补充材料。
