📄 中文摘要
提出了一种新的聚类算法，用于处理纵向数据。这类数据可以被视为由个体组成，每个个体在不同时间点上观察到一个时间依赖变量。不同个体的变量随时间演变的具体方式通常各不相同，但也可能存在共性，即许多个体共享的时间演变特征。该方法的目的是寻找那些其潜在时间依赖变量共享特征的个体聚类。该算法分为两个步骤：第一步将每个个体映射到欧几里得空间中的一个点，坐标由其时间演变特征决定。第二步则基于这些点进行聚类，以识别具有相似时间演变特征的个体群体。
基于特征的轨迹聚类：一种用于纵向数据的聚类算法
📄 English Summary
Introducing Feature-Based Trajectory Clustering, a clustering algorithm for longitudinal data
A new algorithm for clustering longitudinal data is presented. This type of data can be conceptualized as consisting of individuals, each with observations of a time-dependent variable made at various times. The specific way in which this variable evolves over time typically differs from one individual to another, yet there may be commonalities in the characteristic features of time evolution shared by many individuals. The method aims to identify clusters of individuals whose underlying time-dependent variables exhibit such shared features. The algorithm operates in two steps: the first step maps each individual to a point in Euclidean space, with coordinates determined by their time evolution characteristics. The second step performs clustering based on these points to identify groups of individuals with similar time evolution features.