胡贵方《流行病学》20160406-TT-treelets-an adaptive multi-scale basis for sparse unordered data.pdfVIP

NIH Public Access Author Manuscript Ann Appl Stat . Author manuscript; available in PMC 2014 December 02. N Published in final edited form as: I H Ann Appl Stat . 2008 June ; 2(2): 489–493. doi:10.1214/07-AOAS137. - P A A u DISCUSSION OF: TREELETS—AN ADAPTIVE MULTI-SCALE t h o r BASIS FOR SPARSE UNORDERED DATA M a n Catherine Tuglus and Mark J. van der Laan u s University of California, Berkeley c r i p Catherine Tuglus: ctuglus@berkeley.edu; Mark J. van der Laan: laan@berkeley.edu t Abstract We would like to congratulate Lee, Nadler and Wasserman on their contribution to clustering and data reduction methods for high p and low n situations. A composite of clustering and traditional principal components analysis, treelets is an innovative method for multi-resolution analysis of N unordered data. It is an improvement over traditional PCA and an important contribution to I H - clustering methodology. Their paper presents theory and supporting applications addressing the P A two main goals of the treelet method: (1) Uncover the underlying structure of the data and (2) Data A reduction prior to statistical learning methods. We will organize our discussion into two main parts u t h to address their methodology in terms of each of these two goals. We will present and discuss o r treelets in terms of a clustering algorithm and an improvement over traditional PCA. We will also M discuss the applicability of treelets to more general data, in particular, the application of treelets to a n microarra