Functional Data Analysis of Tree Data Objects

摘要: Data analysis on non-Euclidean spaces, such as tree can be challenging. The main contribution of this article is establishment a connection between tree-data spaces and the well-developed area functional data (FDA), where objects are curves. This comes through two representation approaches, Dyck path branch length representation. These representations trees in Euclidean enable us to exploit power FDA explore statistical properties objects. A major challenge sparsity branches sample trees. We overcome issue by using tree-pruning technique that focuses important underlying population structures. method parallels scale-space sense it reveals tree-structured over range scales. effectiveness these new approaches demonstrated some novel results obtained analys...