Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 74 of the 74 references it cites were found in a registry or an index.

Benchmark of filter methods for feature selection in high-dimensional gene expression survival data

Open access

Abstract

Feature selection is crucial for the analysis of high-dimensional data, but benchmark studies for data with a survival outcome are rare. We compare 14 filter methods for feature selection based on 11 high-dimensional gene expression survival data sets. The aim is to provide guidance on the choice of filter methods for other researchers and practitioners. We analyze the accuracy of predictive models that employ the features selected by the filter methods. Also, we consider the run time, the number of selected features for fitting models with high predictive accuracy as well as the feature selection stability. We conclude that the simple variance filter outperforms all other considered filter methods. This filter selects the features with the largest variance and does not take into account the survival outcome. Also, we identify the correlation-adjusted regression scores filter as a more elaborate alternative that allows fitting models with similar predictive accuracy. Additionally, we investigate the filter methods based on feature rankings, finding groups of similar filters.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/bib/bbab354

Cites

74 of 74 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Mean OL scores between the filter methods averaged across data sets. The filter methods are ordered based on single linkage clustering.
Figure 1 Mean OL scores between the filter methods averaged across data sets. The filter methods are ordered based on single linkage clustering.CC BY · PubMed Central
Figure 2 Visualization of nested cross-validation.
Figure 2 Visualization of nested cross-validation.CC BY · PubMed Central
Figure 3 Boxplots of the integrated Brier score of the best configurations of the 10 outer cross-validation iterations per data set. Each boxplot represents 10 performance values. Small values indicate a good predictive performance.
Figure 3 Boxplots of the integrated Brier score of the best configurations of the 10 outer cross-validation iterations per data set. Each boxplot represents 10 performance values. Small values indicate a good predictive performance.CC BY · PubMed Central
Figure 4 Number of data sets on which the filter method displayed in the row achieves a lower mean integrated Brier score (corresponding to a better performance) than the filter method displayed in the column. The filter methods are ordered by the sums of the rows, that is, the total number of wins against any of the other filter methods on any of the 11 data sets.
Figure 4 Number of data sets on which the filter method displayed in the row achieves a lower mean integrated Brier score (corresponding to a better performance) than the filter method displayed in the column. The filter methods are ordered by the sums of the rows, that is, the total number of wins against any of the other filter methods on any of the 11 data sets.CC BY · PubMed Central
Figure 5 Relative mean integrated Brier score and relative logarithmic median run time. A symbol indicates the median across the 11 data sets. The lines span from the minimum to the maximum value observed for any of the data sets. Small values are desirable for both criteria.
Figure 5 Relative mean integrated Brier score and relative logarithmic median run time. A symbol indicates the median across the 11 data sets. The lines span from the minimum to the maximum value observed for any of the data sets. Small values are desirable for both criteria.CC BY · PubMed Central
Figure 6 Boxplots of the proportion of selected features of the best configurations of the 10 outer cross-validation iterations for all data sets. Each boxplot represents 110 proportion values. The mean proportions of selected features are indicated by numbers.
Figure 6 Boxplots of the proportion of selected features of the best configurations of the 10 outer cross-validation iterations for all data sets. Each boxplot represents 110 proportion values. The mean proportions of selected features are indicated by numbers.CC BY · PubMed Central
Figure 7 Boxplots of the feature selection stability of the best configurations of the 10 outer cross-validation iterations for all data sets. Each boxplot represents 110 stability values assessed with the stability measure SMA-Count. Large values are desirable.
Figure 7 Boxplots of the feature selection stability of the best configurations of the 10 outer cross-validation iterations for all data sets. Each boxplot represents 110 stability values assessed with the stability measure SMA-Count. Large values are desirable.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 9

  1. Kittler, 1978, Feature Set Search Algorithms, 41
  2. Tang, 2014, Feature Selection for Classification: A Review, pages 37–64
  3. Bommert, 2020, Integration of Feature Selection Stability in Model Fitting
  4. Kursa, 2020, praznik: Tools for Information-Based Feature Selection
  5. Yang, 1999, Data visualization and feature selection: New algorithms for nongaussian data, 687
  6. Meyer, 2006, On the use of variable complementarity for feature selection in cancer classification, 91
  7. Sonabend, 2020, mlr3proba: Probabilistic Supervised Learning for ‘mlr3’
  8. Simon, 2011, Regularization paths for Cox’s proportional hazards model via coordinate descent, J Stat Softw, 39, 1, 10.18637/jss.v039.i05 · DOI · Paper page
  9. Yang, 2020, OrderedList: Similarities of Ordered Gene Lists

Cited by

See all 182

Metadata

Authors
Bommert, A., Welchowski, T., Schmid, M., & Rahnenführer, J.
Affiliations
  • TU Dortmund University (DE)
  • University of Bonn (DE)
Corresponding author
Bommert, Andrea
Venue
Briefings in Bioinformatics
Published
2021-09-08
Volume
23
Issue
1
ISSN
  • 1467-5463 (ISSN-L)
  • 1477-4054
OpenAlex
W3196757689
Cited by
182
References
74
Citations by year
2026: 38 · 2025: 34 · 2024: 43 · 2023: 37 · 2022: 30
FWCI
8.63
Type
journal article
Language
en
Open access
Locations
  • Briefings in Bioinformatics (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • Deutsche Forschungsgemeinschaft (LiDO3, 271512359, SFB876, SCHM 2966/2-1)
  • Technische Universität Dortmund (SFB 876, 271512359)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
Keywords
feature selection, filter methods, gene expression data, survival analysis, high-dimensional data, benchmarking, feature ranking, prediction accuracy, feature selection stability
MeSH terms
Algorithms, Gene Expression

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.