clusterMLD: An Efficient Hierarchical Clustering Method for Multivariate Longitudinal Data

Longitudinal data clustering is challenging because the grouping has to account for similarity of individual trajectories in presence sparse and irregular times observation. This paper puts forward a hierarchical agglomerative method based on dissimilarity metric that quantifies cost merging two distinct groups curves, which are depicted by B-splines repeatedly measured data. Extensive simulations show proposed superior performance determining number clusters, classifying individuals into correct computational efficiency. Importantly, not only suitable multivariate longitudinal with measurements but also intensely functional Towards this end, we provide an R package implementation such analyses. To illustrate use method, large clinical sets from real-world studies analyzed.