Benchmark for filter methods for feature selection in high-dimensional classification data

Bommert, Andrea; Sun, Xudong; Bischl, Bernd; Rahnenführer, Jörg; Lang, Michel

doi:10.1016/j.csda.2019.106839

Cited by 505 publications

(290 citation statements)

References 45 publications

Supporting

Mentioning

288

Contrasting

Order By: Relevance

“…This value is then used as a measure for the discriminative power of the respective feature. Further filter methods implemented in this paper were chosen according to the recommendations of a recent benchmark study by Bommert et al [10]. They compared 22 filter methods on 16 high-dimensional data sets with respect to predictive performance and runtime.…”

Section: Implementation Of Baseline Filter Methodsmentioning

confidence: 99%

“…The former two methods use an entropy based feature evaluation, while the latter one assesses the node impurity of random forests. For further information on these methods see Brown et al [14], Izenmann et al [15], or Bommert et al [10]. All analyzed filter methods are used with their implementation within the R package mlr [16].…”

Section: Implementation Of Baseline Filter Methodsmentioning

confidence: 99%

“…Besides the greedy approach, we implemented two adaptation strategies for genetic algorithms [8]. As baseline approaches for all of these methods we implemented an unadapted greedy forward feature selection and a simple cost-constraining strategy for four filter methods [9] suggested in Bommert et al [10].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Cost-Constrained feature selection in binary classification: adaptations for greedy forward selection and genetic algorithms

et al. 2020

Self Cite

View full text Add to dashboard Cite

Background: With modern methods in biotechnology, the search for biomarkers has advanced to a challenging statistical task exploring high dimensional data sets. Feature selection is a widely researched preprocessing step to handle huge numbers of biomarker candidates and has special importance for the analysis of biomedical data. Such data sets often include many input features not related to the diagnostic or therapeutic target variable. A less researched, but also relevant aspect for medical applications are costs of different biomarker candidates. These costs are often financial costs, but can also refer to other aspects, for example the decision between a painful biopsy marker and a simple urine test. In this paper, we propose extensions to two feature selection methods to control the total amount of such costs: greedy forward selection and genetic algorithms. In comprehensive simulation studies of binary classification tasks, we compare the predictive performance, the run-time and the detection rate of relevant features for the new proposed methods and five baseline alternatives to handle budget constraints. Results: In simulations with a predefined budget constraint, our proposed methods outperform the baseline alternatives, with just minor differences between them. Only in the scenario without an actual budget constraint, our adapted greedy forward selection approach showed a clear drop in performance compared to the other methods. However, introducing a hyperparameter to adapt the benefit-cost trade-off in this method could overcome this weakness. Conclusions: In feature cost scenarios, where a total budget has to be met, common feature selection algorithms are often not suitable to identify well performing subsets for a modelling task. Adaptations of these algorithms such as the ones proposed in this paper can help to tackle this problem.

show abstract

Section: Implementation Of Baseline Filter Methodsmentioning

confidence: 99%

Section: Implementation Of Baseline Filter Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Cost-Constrained feature selection in binary classification: adaptations for greedy forward selection and genetic algorithms

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…is as defined in Eq. (5). As there are datasets with millions of features we seek an algorithm to select f t with linear complexity.…”

Section: Iterative Feature Selectionmentioning

confidence: 99%

“…Among the three main feature selection methods, filter methods are preferred to wrapper and embedded methods in applications where the computational efficiency, classifier independence, simplicity, ease of use and the stability of the results are required. Therefore, filter feature selection remains an interesting topic in many recent research areas such as biomarker identification for cancer prediction and drugs discovery, text classification and predicting defective software [3][4][5]10,11,16,18] and has growing interest in big data applications [19]; according to the Google Scholar search results, the number of research papers published related to filter methods in year 2018 is ∼1,800 of which ∼170 are in gene selection area.…”

Section: Introductionmentioning

confidence: 99%

A Framework for Feature Selection to Exploit Feature Group Structures

Perera

Chan

Karunasekera

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Filter feature selection methods play an important role in machine learning tasks when low computational costs, classifier independence or simplicity is important. Existing filter methods predominantly focus only on the input data and do not take advantage of the external sources of correlations within feature groups to improve the classification accuracy. We propose a framework which facilitates supervised filter feature selection methods to exploit feature group information from external sources of knowledge and use this framework to incorporate feature group information into minimum Redundancy Maximum Relevance (mRMR) algorithm, resulting in GroupMRMR algorithm. We show that GroupMRMR achieves high accuracy gains over mRMR (up to ∼35%) and other popular filter methods (up to ∼50%). GroupMRMR has same computational complexity as that of mRMR, therefore, does not incur additional computational costs. Proposed method has many real world applications, particularly the ones that use genomic, text and image data whose features demonstrate strong group structures.

show abstract