Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
A new nonparametric screening method for ultrahigh-dimensional survival data
Liu Y., Zhang J., Zhao X. Computational Statistics & Data Analysis119 (C):74-85,2018.Type:Article
Date Reviewed: May 31 2018

Groundbreaking effective algorithms for overcoming the discrepancies in the existing screening procedures for ultrahigh-dimensional medical survival data are long overdue. Liu et al. offer a nonparametric Kolmogorov-Smirnov (K-S) test statistic for understanding the ultrahigh-dimensional survival patterns in the historical genomic and health datasets of patients. They propose a model-free screening procedure that incorporates alternative covariate variables of patient data into a data analysis for exploring medicinal treatments in areas such as cancer.

The authors convincingly present algorithms for independently applying the K-S test statistic, and well-known survival data distributions for screening patient data over time. The exponential variation in alternative patient data is used to illuminate the steadfast screening characteristics of the bid K-S-based procedure for exploring patterns in ultrahigh-dimensional censored data. The authors perform theoretical and practical simulation experiments to illustrate the effectiveness of the proposed K-S statistic and the new procedure for use in medical screenings of the ultrahigh-dimensional survival data of patients. The theoretical simulation experiments show reliable data modeling results that mimic hazards, nonlinear interaction, and time series models.

In an experimental study, the proposed nonparametric procedure was reliably used to identify the genes germane to the survival of cancer. The nonparametric method allows for categorical and continuous variables to be incorporated as covariates in the screening of ultrahigh-dimensional patient data, and to reliably forecast the survival of patients in medical treatments using differently scaled historical records. All computational statisticians ought to read and comment on the theoretical proofs of the lemmas underlying the K-S-based procedures outlined in this paper.

Reviewer:  Amos Olagunju Review #: CR146055 (1808-0439)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Nonparametric Statistics (G.3 ... )
 
 
Medical Information Systems (J.3 ... )
 
 
Learning (I.2.6 )
 
 
Pattern Recognition (I.5 )
 
Would you recommend this review?
yes
no
Other reviews under "Nonparametric Statistics": Date
Nonparametric econometrics
Pagan A., Ullah A., Cambridge University Press, New York, NY, 1999. Type: Book (9780521355643)
Aug 1 1999
Baysian nonparametrics via neural networks
Lee H., Society for Industrial & Applied Mathematics., 2004. Type: Book (9780898715637)
Jan 18 2005
On projection-based tests for directional and compositional data
Cuesta-Albertos J., Cuevas A., Fraiman R. Statistics and Computing 19(4): 367-380, 2009. Type: Article
Mar 22 2010
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy